3 components of a research journal

• UNO Criss Library

Social Work Research Guide

What is a research journal.

• Find Articles
• Find E-Books and Books

Reading an Academic Article

• Free Online Resources
• Reference and Writing
• Citation Help This link opens in a new window

Anatomy of a Scholarly Article

TIP: When possible, keep your research question(s) in mind when reading scholarly articles. It will help you to focus your reading.

Title : Generally are straightforward and describe what the article is about. Titles often include relevant key words.

Abstract : A summary of the author(s)'s research findings and tells what to expect when you read the full article. It is often a good idea to read the abstract first, in order to determine if you should even bother reading the whole article.

Discussion and Conclusion : Read these after the Abstract (even though they come at the end of the article). These sections can help you see if this article will meet your research needs. If you don’t think that it will, set it aside.

Introduction : Describes the topic or problem researched. The authors will present the thesis of their argument or the goal of their research.

Literature Review : May be included in the introduction or as its own separate section. Here you see where the author(s) enter the conversation on this topic. That is to say, what related research has come before, and how do they hope to advance the discussion with their current research?

Methods : This section explains how the study worked. In this section, you often learn who and how many participated in the study and what they were asked to do. You will need to think critically about the methods and whether or not they make sense given the research question.

Results : Here you will often find numbers and tables. If you aren't an expert at statistics this section may be difficult to grasp. However you should attempt to understand if the results seem reasonable given the methods.

Works Cited (also be called References or Bibliography ): This section comprises the author(s)’s sources. Always be sure to scroll through them. Good research usually cites many different kinds of sources (books, journal articles, etc.). As you read the Works Cited page, be sure to look for sources that look like they will help you to answer your own research question.

Adapted from http://library.hunter.cuny.edu/research-toolkit/how-do-i-read-stuff/anatomy-of-a-scholarly-article

A research journal is a periodical that contains articles written by experts in a particular field of study who report the results of research in that field. The articles are intended to be read by other experts or students of the field, and they are typically much more sophisticated and advanced than the articles found in general magazines. This guide offers some tips to help distinguish scholarly journals from other periodicals.

CHARACTERISTICS OF RESEARCH JOURNALS

PURPOSE : Research journals communicate the results of research in the field of study covered by the journal. Research articles reflect a systematic and thorough study of a single topic, often involving experiments or surveys. Research journals may also publish review articles and book reviews that summarize the current state of knowledge on a topic.

APPEARANCE : Research journals lack the slick advertising, classified ads, coupons, etc., found in popular magazines. Articles are often printed one column to a page, as in books, and there are often graphs, tables, or charts referring to specific points in the articles.

AUTHORITY : Research articles are written by the person(s) who did the research being reported. When more than two authors are listed for a single article, the first author listed is often the primary researcher who coordinated or supervised the work done by the other authors. The most highly‑regarded scholarly journals are typically those sponsored by professional associations, such as the American Psychological Association or the American Chemical Society.

VALIDITY AND RELIABILITY : Articles submitted to research journals are evaluated by an editorial board and other experts before they are accepted for publication. This evaluation, called peer review, is designed to ensure that the articles published are based on solid research that meets the normal standards of the field of study covered by the journal. Professors sometimes use the term "refereed" to describe peer-reviewed journals.

WRITING STYLE : Articles in research journals usually contain an advanced vocabulary, since the authors use the technical language or jargon of their field of study. The authors assume that the reader already possesses a basic understanding of the field of study.

REFERENCES : The authors of research articles always indicate the sources of their information. These references are usually listed at the end of an article, but they may appear in the form of footnotes, endnotes, or a bibliography.

PERIODICALS THAT ARE NOT RESEARCH JOURNALS

POPULAR MAGAZINES : These are periodicals that one typically finds at grocery stores, airport newsstands, or bookstores at a shopping mall. Popular magazines are designed to appeal to a broad audience, and they usually contain relatively brief articles written in a readable, non‑technical language.

Examples include: Car and Driver , Cosmopolitan , Esquire , Essence , Gourmet , Life , People Weekly , Readers' Digest , Rolling Stone , Sports Illustrated , Vanity Fair , and Vogue .

NEWS MAGAZINES : These periodicals, which are usually issued weekly, provide information on topics of current interest, but their articles seldom have the depth or authority of scholarly articles.

Examples include: Newsweek , Time , U.S. News and World Report .

OPINION MAGAZINES : These periodicals contain articles aimed at an educated audience interested in keeping up with current events or ideas, especially those pertaining to topical issues. Very often their articles are written from a particular political, economic, or social point of view.

Examples include: Catholic World , Christianity Today , Commentary , Ms. , The Militant , Mother Jones , The Nation , National Review , The New Republic , The Progressive , and World Marxist Review .

TRADE MAGAZINES : People who need to keep up with developments in a particular industry or occupation read these magazines. Many trade magazines publish one or more special issues each year that focus on industry statistics, directory lists, or new product announcements.

Examples include: Beverage World , Progressive Grocer , Quick Frozen Foods International , Rubber World , Sales and Marketing Management , Skiing Trade News , and Stores .

Literature Reviews

• Literature Review Guide General information on how to organize and write a literature review.
• The Literature Review: A Few Tips On Conducting It Contains two sets of questions to help students review articles, and to review their own literature reviews.
• << Previous: Find E-Books and Books
• Next: Statistics >>
• Last Updated: Mar 8, 2024 11:27 AM
• URL: https://libguides.unomaha.edu/social_work

• Research Guides

BSCI 1510L Literature and Stats Guide: 3.2 Components of a scientific paper

• 1 What is a scientific paper?
• 2 Referencing and accessing papers
• 2.1 Literature Cited
• 2.2 Accessing Scientific Papers
• 2.3 Traversing the web of citations
• 2.4 Keyword Searches
• 3 Style of scientific writing
• 3.1 Specific details regarding scientific writing

3.2 Components of a scientific paper

• 4 For further information
• Appendix A: Calculation Final Concentrations
• 1 Formulas in Excel
• 2 Basic operations in Excel
• 3 Measurement and Variation
• 3.1 Describing Quantities and Their Variation
• 3.2 Samples Versus Populations
• 3.3 Calculating Descriptive Statistics using Excel
• 4 Variation and differences
• 5 Differences in Experimental Science
• 5.1 Aside: Commuting to Nashville
• 5.2 P and Detecting Differences in Variable Quantities
• 5.3 Statistical significance
• 5.4 A test for differences of sample means: 95% Confidence Intervals
• 5.5 Error bars in figures
• 5.6 Discussing statistics in your scientific writing
• 6 Scatter plot, trendline, and linear regression
• 7 The t-test of Means
• 8 Paired t-test
• 9 Two-Tailed and One-Tailed Tests
• 10 Variation on t-tests: ANOVA
• 11 Reporting the Results of a Statistical Test
• 12 Summary of statistical tests
• 1 Objectives
• 2 Project timeline
• 3 Background
• 4 Previous work in the BSCI 111 class
• 5 General notes about the project
• 6 About the paper
• 7 References

Nearly all journal articles are divided into the following major sections: abstract, introduction, methods, results, discussion, and references.  Usually the sections are labeled as such, although often the introduction (and sometimes the abstract) is not labeled.  Sometimes alternative section titles are used.  The abstract is sometimes called the "summary", the methods are sometimes called "materials and methods", and the discussion is sometimes called "conclusions".   Some journals also include the minor sections of "key words" following the abstract, and "acknowledgments" following the discussion.  In some journals, the sections may be divided into subsections that are given descriptive titles.  However, the general division into the six major sections is nearly universal.

3.2.1 Abstract

The abstract is a short summary (150-200 words or less) of the important points of the paper.  It does not generally include background information.  There may be a very brief statement of the rationale for conducting the study.  It describes what was done, but without details.  It also describes the results in a summarized way that usually includes whether or not the statistical tests were significant.  It usually concludes with a brief statement of the importance of the results.  Abstracts do not include references.  When writing a paper, the abstract is always the last part to be written.

The purpose of the abstract is to allow potential readers of a paper to find out the important points of the paper without having to actually read the paper.  It should be a self-contained unit capable of being understood without the benefit of the text of the article . It essentially serves as an "advertisement" for the paper that readers use to determine whether or not they actually want to wade through the entire paper or not.  Abstracts are generally freely available in electronic form and are often presented in the results of an electronic search.  If searchers do not have electronic access to the journal in which the article is published, the abstract is the only means that they have to decide whether to go through the effort (going to the library to look up the paper journal, requesting a reprint from the author, buying a copy of the article from a service, requesting the article by Interlibrary Loan) of acquiring the article.  Therefore it is important that the abstract accurately and succinctly presents the most important information in the article.

3.2.2 Introduction

The introduction provides the background information necessary to understand why the described experiment was conducted.  The introduction should describe previous research on the topic that has led to the unanswered questions being addressed by the experiment and should cite important previous papers that form the background for the experiment.  The introduction should also state in an organized fashion the goals of the research, i.e. the particular, specific questions that will be tested in the experiments.  There should be a one-to-one correspondence between questions raised in the introduction and points discussed in the conclusion section of the paper.  In other words, do not raise questions in the introduction unless you are going to have some kind of answer to the question that you intend to discuss at the end of the paper. 

You may have been told that every paper must have a hypothesis that can be clearly stated.  That is often true, but not always.  If your experiment involves a manipulation which tests a specific hypothesis, then you should clearly state that hypothesis.  On the other hand, if your experiment was primarily exploratory, descriptive, or measurative, then you probably did not have an a priori hypothesis, so don't pretend that you did and make one up.  (See the discussion in the introduction to Experiment 4 for more on this.)  If you state a hypothesis in the introduction, it should be a general hypothesis and not a null or alternative hypothesis for a statistical test.  If it is necessary to explain how a statistical test will help you evaluate your general hypothesis, explain that in the methods section. 

A good introduction should be fairly heavy with citations.  This indicates to the reader that the authors are informed about previous work on the topic and are not working in a vacuum.  Citations also provide jumping-off points to allow the reader to explore other tangents to the subject that are not directly addressed in the paper.  If the paper supports or refutes previous work, readers can look up the citations and make a comparison for themselves. 

"Do not get lost in reviewing background information. Remember that the Introduction is meant to introduce the reader to your research, not summarize and evaluate all past literature on the subject (which is the purpose of a review paper). Many of the other studies you may be tempted to discuss in your Introduction are better saved for the Discussion, where they become a powerful tool for comparing and interpreting your results. Include only enough background information to allow your reader to understand why you are asking the questions you are and why your hyptheses are reasonable ones. Often, a brief explanation of the theory involved is sufficient. …

Write this section in the past or present tense, never in the future. " (Steingraber et al. 1985)

3.2.3 Methods (taken verbatim from Steingraber et al. 1985)

The function of this section is to describe all experimental procedures, including controls. The description should be complete enough to enable someone else to repeat your work. If there is more than one part to the experiment, it is a good idea to describe your methods and present your results in the same order in each section. This may not be the same order in which the experiments were performed -it is up to you to decide what order of presentation will make the most sense to your reader.

1. Explain why each procedure was done, i.e., what variable were you measuring and why? Example:

Difficult to understand : First, I removed the frog muscle and then I poured Ringer’s solution on it. Next, I attached it to the kymograph.

Improved: I removed the frog muscle and poured Ringer’s solution on it to prevent it from drying out. I then attached the muscle to the kymograph in order to determine the minimum voltage required for contraction.

2. Experimental procedures and results are narrated in the past tense (what you did, what you found, etc.) whereas conclusions from your results are given in the present tense.

3. Mathematical equations and statistical tests are considered mathematical methods and should be described in this section along with the actual experimental work.

4. Use active rather than passive voice when possible.  [Note: see Section 3.1.4 for more about this.]  Always use the singular "I" rather than the plural "we" when you are the only author of the paper.  Throughout the paper, avoid contractions, e.g. did not vs. didn’t.

5. If any of your methods is fully described in a previous publication (yours or someone else’s), you can cite that instead of describing the procedure again.

Example: The chromosomes were counted at meiosis in the anthers with the standard acetocarmine technique of Snow (1955).

3.2.4 Results (with excerpts from Steingraber et al. 1985)

The function of this section is to summarize general trends in the data without comment, bias, or interpretation. The results of statistical tests applied to your data are reported in this section although conclusions about your original hypotheses are saved for the Discussion section.

Tables and figures should be used when they are a more efficient way to convey information than verbal description. They must be independent units, accompanied by explanatory captions that allow them to be understood by someone who has not read the text. Do not repeat in the text the information in tables and figures, but do cite them, with a summary statement when that is appropriate.  Example:

Incorrect: The results are given in Figure 1.

Correct: Temperature was directly proportional to metabolic rate (Fig. 1).

Please note that the entire word "Figure" is almost never written in an article.  It is nearly always abbreviated as "Fig." and capitalized.  Tables are cited in the same way, although Table is not abbreviated.

Whenever possible, use a figure instead of a table. Relationships between numbers are more readily grasped when they are presented graphically rather than as columns in a table.

Data may be presented in figures and tables, but this may not substitute for a verbal summary of the findings. The text should be understandable by someone who has not seen your figures and tables.

1. All results should be presented, including those that do not support the hypothesis.

2. Statements made in the text must be supported by the results contained in figures and tables.

3. The results of statistical tests can be presented in parentheses following a verbal description.

Example: Fruit size was significantly greater in trees growing alone (t = 3.65, df = 2, p < 0.05).

Simple results of statistical tests may be reported in the text as shown in the preceding example.  The results of multiple tests may be reported in a table if that increases clarity. (See Section 11 of the Statistics Manual for more details about reporting the results of statistical tests.)  It is not necessary to provide a citation for a simple t-test of means, paired t-test, or linear regression.  If you use other tests, you should cite the text or reference you followed to do the test.  In your materials and methods section, you should report how you did the test (e.g. using the statistical analysis package of Excel). 

It is NEVER appropriate to simply paste the results from statistical software into the results section of your paper.  The output generally reports more information than is required and it is not in an appropriate format for a paper.

3.2.4.1 Tables

• Do not repeat information in a table that you are depicting in a graph or histogram; include a table only if it presents new information.
• It is easier to compare numbers by reading down a column rather than across a row. Therefore, list sets of data you want your reader to compare in vertical form.
• Provide each table with a number (Table 1, Table 2, etc.) and a title. The numbered title is placed above the table .
• Please see Section 11 of the Excel Reference and Statistics Manual for further information on reporting the results of statistical tests.

3.2.4.2. Figures

• These comprise graphs, histograms, and illustrations, both drawings and photographs. Provide each figure with a number (Fig. 1, Fig. 2, etc.) and a caption (or "legend") that explains what the figure shows. The numbered caption is placed below the figure .  Figure legend = Figure caption.
• Figures submitted for publication must be "photo ready," i.e., they will appear just as you submit them, or photographically reduced. Therefore, when you graduate from student papers to publishable manuscripts, you must learn to prepare figures that will not embarrass you. At the present time, virtually all journals require manuscripts to be submitted electronically and it is generally assumed that all graphs and maps will be created using software rather than being created by hand.  Nearly all journals have specific guidelines for the file types, resolution, and physical widths required for figures.  Only in a few cases (e.g. sketched diagrams) would figures still be created by hand using ink and those figures would be scanned and labeled using graphics software.  Proportions must be the same as those of the page in the journal to which the paper will be submitted. 
• Graphs and Histograms: Both can be used to compare two variables. However, graphs show continuous change, whereas histograms show discrete variables only.  You can compare groups of data by plotting two or even three lines on one graph, but avoid cluttered graphs that are hard to read, and do not plot unrelated trends on the same graph. For both graphs, and histograms, plot the independent variable on the horizontal (x) axis and the dependent variable on the vertical (y) axis. Label both axes, including units of measurement except in the few cases where variables are unitless, such as absorbance.
• Drawings and Photographs: These are used to illustrate organisms, experimental apparatus, models of structures, cellular and subcellular structure, and results of procedures like electrophoresis. Preparing such figures well is a lot of work and can be very expensive, so each figure must add enough to justify its preparation and publication, but good figures can greatly enhance a professional article, as your reading in biological journals has already shown.

3.2.5 Discussion (taken from Steingraber et al. 1985)

The function of this section is to analyze the data and relate them to other studies. To "analyze" means to evaluate the meaning of your results in terms of the original question or hypothesis and point out their biological significance.

1. The Discussion should contain at least:

• the relationship between the results and the original hypothesis, i.e., whether they support the hypothesis, or cause it to be rejected or modified
• an integration of your results with those of previous studies in order to arrive at explanations for the observed phenomena
• possible explanations for unexpected results and observations, phrased as hypotheses that can be tested by realistic experimental procedures, which you should describe

2. Trends that are not statistically significant can still be discussed if they are suggestive or interesting, but cannot be made the basis for conclusions as if they were significant.

3. Avoid redundancy between the Results and the Discussion section. Do not repeat detailed descriptions of the data and results in the Discussion. In some journals, Results and Discussions are joined in a single section, in order to permit a single integrated treatment with minimal repetition. This is more appropriate for short, simple articles than for longer, more complicated ones.

4. End the Discussion with a summary of the principal points you want the reader to remember. This is also the appropriate place to propose specific further study if that will serve some purpose, but do not end with the tired cliché that "this problem needs more study." All problems in biology need more study. Do not close on what you wish you had done, rather finish stating your conclusions and contributions.

3.2.6 Title

The title of the paper should be the last thing that you write.  That is because it should distill the essence of the paper even more than the abstract (the next to last thing that you write). 

The title should contain three elements:

1. the name of the organism studied;

2. the particular aspect or system studied;

3. the variable(s) manipulated.

Do not be afraid to be grammatically creative. Here are some variations on a theme, all suitable as titles:

THE EFFECT OF TEMPERATURE ON GERMINATION OF ZEA MAYS

DOES TEMPERATURE AFFECT GERMINATION OF ZEA MAYS?

TEMPERATURE AND ZEA MAYS GERMINATION: IMPLICATIONS FOR AGRICULTURE

Sometimes it is possible to include the principal result or conclusion in the title:

HIGH TEMPERATURES REDUCE GERMINATION OF ZEA MAYS

Note for the BSCI 1510L class: to make your paper look more like a real paper, you can list all of the other group members as co-authors.  However, if you do that, you should list you name first so that we know that you wrote it.

3.2.7 Literature Cited

Please refer to section 2.1 of this guide.

• << Previous: 3.1 Specific details regarding scientific writing
• Next: 4 For further information >>
• Last Updated: Apr 19, 2023 2:37 PM
• URL: https://researchguides.library.vanderbilt.edu/bsci1510L

Research Basics Tutorial: Parts of a Scholarly Article

• Tutorial Menu
• College-Level Research
• What Do Researchers Do?
• Parts of a Scholarly Article
• How to Search
• Finding Books & Articles
• Evaluating Sources
• Literature Reviews
• Web vs. Google Scholar
• Citing Sources

Qualitative Methods

Key components of a scholarly article.

Components of a Research Article

The title may include terms like “outcomes,” “effects,” “treatments,” and “reactions” that indicate the article deals with research. Example:

Human touch effectively and safely reduces pain in the newborn intensive care unit

Provides the author(s) name(s) and publication information. Example:

Herrington, C. J., & Chiodo, L. M. (2014). Human touch effectively and safely reduces pain in the newborn intensive care unit. Pain Management Nursing , 15 (1), 107-115. doi:10.1016/j.pmn.2012.06.007

Used by readers to quickly evaluate the overall article content. Introduces topic and specific research question, usually provides statement regarding methodology and a general statement about the results and findings.

This was a feasibility pilot study to evaluate the efficacy of the nonpharmacologic pain management technique of gentle human touch (GHT) in reducing pain response to heel stick in premature infants in the neonatal intensive care unit (NICU). Eleven premature infants ranging from 27 to 34 weeks’ gestational age, in a level III NICU in a teaching hospital, were recruited and randomized to order of treatment in this repeated-measures crossover-design experiment. Containment with GHT during heel stick was compared with traditional nursery care (side lying and “nested” in an incubator). Heart rate, respiratory rate, oxygen saturation, and cry were measured continuously beginning at baseline and continuing through heel warming, heel stick, and recovery following the heel stick. Infants who did not receive GHT had decreased respiration, increased heart rate, and increased cry time during the heel stick... No significant differences were noted in oxygen saturation in either group. GHT is a simple nonpharmacologic therapy that can be used by nurses and families to reduce pain of heel stick in premature infants in the NICU.

Introduction

Introduces the broad overall topic and provides basic background information.  Then narrows down to the specific research question relating to the topic. Provides the purpose and focus for the rest of the article and sets the justification for the research. Sometimes this includes a literature review that describes past important research and relates it specifically to the research question.

Background: Nearly 13% of all pregnancies result in premature birth (infants born before 37 weeks’ completed gestation) every year in the U.S. ( Martin, Hamilton, Sutton, Ventura, Mathews, & Osterman, 2010 ). It is estimated that 50%–70% of infants born prematurely develop neurobehavioral deficits/delays that are often undiagnosed until preschool and early school age…. Although multiple mechanisms affect overall neurobehavioral development in these infants, increased scientific attention has focused on the detrimental effects of minor repetitive pain exposure in the newborn intensive care unit (NICU) ( Fitzgerald & Walker, 2009 ; Grunau, 2002 ; Hack, Klein, & Taylor, 1995 ). . . .

Study Aim: The present study was designed to test the efficacy of gentle human touch in reducing pain response in premature infants undergoing heel stick for medically indicated blood sampling compared with standard nursery care of positioning with nonhuman confinement using “blanket nesting.”

Describes the research design and methodology used to complete the study.  Includes the sample of who was studied and sample size.

Study Design: This was an experimental pilot feasibility study using a repeated-measures crossover study design. Pain response was measured around two medically indicated heel sticks for blood sampling. Each infant received one heel stick with GHT intervention and one heel stick without GHT. Infants served as their own controls with random assignment to order of treatment (GHT vs. no GHT) in blocks of four to maximize study power. . . .

Sample: Sample size was determined using feasibility considerations of the number of premature infants treated in the unit where the study was conducted. . . .

Results of the analysis are presented that are directly related to the research or problem.

Data Analysis: Descriptive statistics were used to analyze the demographic data and evaluate distributions, measures of central tendency, and outcome variable variability (HR, RR, SaO 2 , and cry). In SPSS (version 20), repeated-measures analysis of variance was used to analyze the effects of intervention versus no intervention. . . .

Discussion / Conclusion

This section should be a discussion of the results, and the implications on the field. The research question should be answered and validated by the interpretation of the results. This section could also discuss how results relate to previous research, any cautions about the findings, and potential for future research.

Discussion: The data reported here provide evidence of the ability of gentle human touch to reduce pain response in premature infants undergoing heel stick for medically indicated blood sampling compared with standard nursery care of positioning with nonhuman confinement using “blanket nesting. . . .”

Study Limitations: There are study limitations that need to be considered as the findings are interpreted. Although statistical significance was noted in several between-group comparisons, the sample was small, thus threatening study validity. However, the clinical significance is important. . . .

Recommendations for Future Research: …Future research should examine the efficacy of GHT to reduce pain response over the duration of the NICU stay and the potential for sensitization to the GHT when used consistently for pain reduction. . . .

Summary: This study presents new evidence supporting the feasibility and effectiveness of gentle human touch for relief of pain of heel stick in the NICU. GHT is a quick and easy intervention that can be provided by nurses to reduce pain.

This section should be an alphabetized list of all the academic sources of information utilized in the paper.

Axelin, A., Salantera, S., Lehtonen, L. (2006). Facilitated tucking by parents’ in pain management of preterm infants—A randomized crossover trial. Early Human Development, 82 (4) (2006), pp. 241–247.

Fitzgerald, M., Walker, S.M. (2009). Infant pain management: A developmental neurobiological approach . Nature Clinical Practice. Neurology, 5 (1) pp. 35–50.

A Few Basic Types of Scholarly Articles

Peer review isn't a tough concept. It just means that the article was reviewed by scholars and meets certain standards with regards to a publication or a discipline. These articles are typically written by professors or specialists. A great clue that something is peer reviewed or scholarly: The article contains a list of references or cites.

Qualitative research seeks to understanding some aspect of social life, and its methods (usually) generate words, rather than numbers, as data for analysis. These research methods seek to understand the experiences and attitudes of the people being studied. They answer questions about the “what,” “how, “ or “why” of a phenomenon rather than “how many” or “how much,” which are answered by quantitative methods.

• << Previous: What Do Researchers Do?
• Next: How to Search >>
• Last Updated: Oct 18, 2022 3:39 PM
• URL: https://libguides.moval.edu/infolit-tutorial

Structure of a Research Paper

Structure of a Research Paper: IMRaD Format

I. The Title Page

• Title: Tells the reader what to expect in the paper.
• Author(s): Most papers are written by one or two primary authors. The remaining authors have reviewed the work and/or aided in study design or data analysis (International Committee of Medical Editors, 1997). Check the Instructions to Authors for the target journal for specifics about authorship.
• Keywords [according to the journal]
• Corresponding Author: Full name and affiliation for the primary contact author for persons who have questions about the research.
• Financial & Equipment Support [if needed]: Specific information about organizations, agencies, or companies that supported the research.
• Conflicts of Interest [if needed]: List and explain any conflicts of interest.

II. Abstract: “Structured abstract” has become the standard for research papers (introduction, objective, methods, results and conclusions), while reviews, case reports and other articles have non-structured abstracts. The abstract should be a summary/synopsis of the paper.

III. Introduction: The “why did you do the study”; setting the scene or laying the foundation or background for the paper.

IV. Methods: The “how did you do the study.” Describe the --

• Context and setting of the study
• Specify the study design
• Population (patients, etc. if applicable)
• Sampling strategy
• Intervention (if applicable)
• Identify the main study variables
• Data collection instruments and procedures
• Outline analysis methods

V. Results: The “what did you find” --

• Report on data collection and/or recruitment
• Participants (demographic, clinical condition, etc.)
• Present key findings with respect to the central research question
• Secondary findings (secondary outcomes, subgroup analyses, etc.)

VI. Discussion: Place for interpreting the results

• Main findings of the study
• Discuss the main results with reference to previous research
• Policy and practice implications of the results
• Strengths and limitations of the study

VII. Conclusions: [occasionally optional or not required]. Do not reiterate the data or discussion. Can state hunches, inferences or speculations. Offer perspectives for future work.

VIII. Acknowledgements: Names people who contributed to the work, but did not contribute sufficiently to earn authorship. You must have permission from any individuals mentioned in the acknowledgements sections. 

IX. References:  Complete citations for any articles or other materials referenced in the text of the article.

• IMRD Cheatsheet (Carnegie Mellon) pdf.
• Adewasi, D. (2021 June 14).  What Is IMRaD? IMRaD Format in Simple Terms! . Scientific-editing.info. 
• Nair, P.K.R., Nair, V.D. (2014). Organization of a Research Paper: The IMRAD Format. In: Scientific Writing and Communication in Agriculture and Natural Resources. Springer, Cham. https://doi.org/10.1007/978-3-319-03101-9_2
• Sollaci, L. B., & Pereira, M. G. (2004). The introduction, methods, results, and discussion (IMRAD) structure: a fifty-year survey.   Journal of the Medical Library Association : JMLA ,  92 (3), 364–367.
• Cuschieri, S., Grech, V., & Savona-Ventura, C. (2019). WASP (Write a Scientific Paper): Structuring a scientific paper.   Early human development ,  128 , 114–117. https://doi.org/10.1016/j.earlhumdev.2018.09.011

What is Research?: Parts of a Research Article

• The Truth about Research
• Research Steps
• Evaluating Sources
• Parts of a Research Article

While each article is different, here are some common pieces you'll see in many of them...

• The title of the article should give you some clues as to the topic it addresses.
• The abstract allows readers to quickly review the overall content of the article. It should give you an idea of the topic of the article, while also providing any key details--such as the questions address in the article and the general results of the studies conducted.
• The introduction introduces the general topic and provides some background information, eventually narrowing it down to the specific issues addressed in the article.
• The literature review describes past research on the topic and relates it to the specific topic covered by the article.  Not all articles will have a literature review.
• The methods section addresses the research design and methodology used by the author to come to the conclusions they have in this article.  This gives others the ability to replicate the study.  Not all articles will have this, since there will be many articles that don't involve an actual study.
• The results section presents the results of any studies or analysis that has been conducted.  Not all articles will have this, either.
• The discussion/conclusion addresses the implications or future of the field.  It may also address where future research is needed.
• The list references or bibliography is the alphabetized list of resources used for the article.  The format of the citations is often determined by what that field's preferred format is.  Common citations formats include APA, Chicago, and MLA.  This is a necessity in an article--and it helps you identify more possible resources for your own paper.
• Components of a Research Paper Useful site that goes more in depth on these sections.
• Parts of a Citation A really wonderful site by the Nash Community College Library.
• << Previous: Evaluating Sources
• Last Updated: Jul 20, 2017 9:23 PM
• URL: https://libguides.uno.edu/whatisresearch

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Am J Speech Lang Pathol

Evidence-Based Practice in Speech-Language Pathology: Where Are We Now?

Tamar greenwell.

a Department of Speech, Language, and Hearing Sciences, Purdue University, West Lafayette, IN

c Tamar Greenwell and Bridget Walsh share first authorship.

Bridget Walsh

b Department of Communicative Sciences and Disorders, Michigan State University, East Lansing

In 2004, American Speech-Language-Hearing Association established its position statement on evidence-based practice (EBP). Since 2008, the Council on Academic Accreditation has required accredited graduate education programs in speech-language pathology to incorporate research methodology and EBP principles into their curricula and clinical practicums. Over the past 15 years, access to EBP resources and employer-led EBP training opportunities have increased. The purpose of this study is to provide an update of how increased exposure to EBP principles affects reported use of EBP and perceived barriers to providing EBP in clinical decision making.

Three hundred seventeen speech-language pathologists completed an online questionnaire querying their perceptions about EBP, use of EBP in clinical practice, and perceived barriers to incorporating EBP. Participants' responses were analyzed using descriptive and inferential statistics. We used multiple linear regression to examine whether years of practice, degree, EBP exposure during graduate program and clinical fellowship (CF), EBP career training, and average barrier score predicted EBP use.

Exposure to EBP in graduate school and during the CF, perception of barriers, and EBP career training significantly predicted the use of EBP in clinical practice. Speech-language pathologists identified the three major components of EBP: client preferences, external evidence, and clinical experience as the most frequently turned to sources of EBP. Inadequate time for research and workload/caseload size remain the most significant barriers to EBP implementation. Respondents who indicated time was a barrier were more likely to cite other barriers to implementing EBP. An increase in EBP career training was associated with a decrease in the perception of time as a barrier.

Conclusions

These findings suggest that explicit training in graduate school and during the CF lays a foundation for EBP principles that is shaped through continued learning opportunities. We documented positive attitudes toward EBP and consistent application of the three components of EBP in clinical practice. Nevertheless, long-standing barriers remain. We suggest that accessible, time-saving resources, a consistent process for posing and answering clinical questions, and on the job support and guidance from employers/organizations are essential to implementing clinical practices that are evidence based. The implications of our findings and suggestions for future research to bridge the research-to-practice gap are discussed.

Promoting the understanding and use of evidence in clinical practice through explicit instruction in the classroom and clinic has long been an objective of graduate programs in speech-language pathology. A committee of the American Speech-Language-Hearing Association (ASHA) on evidence-based practice (EBP) was formed in 2004 to review clinical practices in the field at that time. The committee established the following position statement regarding EBP: “An approach in which current, high-quality research evidence is integrated with practitioner expertise and client preferences and values into the process of making clinical decisions” ( ASHA, 2004 ). ASHA's statement reflects the influential position proposed by Sackett et al. (1996) that “Evidence based medicine is the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients” (p. 71). Since 2008, the Council on Academic Accreditation ( CAA, 2014 ) has required accredited graduate education programs in audiology and speech-language pathology to include research methodology and EBP principles into their curricula and clinical practicum. ASHA certification standards in speech-language pathology incorporate three principles into their definition of EBP: EBP is the integration of evidence, for example, from the scientific literature, with clinical expertise and experience, and client preferences. Graduate students aspiring to receive the Certificate of Clinical Competence must demonstrate a proficiency in these skills related to both evaluation and intervention. Familiarity with EBP and access to EBP resources are key components to promoting its use in clinical practice.

After ASHA released its EBP statement in 2004, Zipoli and Kennedy (2005) conducted a survey of 240 speech-language pathologists to explore the perception, use, and perceived barriers of providing EBP. Despite a growing body of EBP literature in other health professions, Zipoli and Kennedy's seminal study was among the earliest in the field of speech-language pathology. They found that greater exposure to EBP principles and research during participants' clinical fellowships (CFs) translated into greater reported use of EBP. Speech-language pathologists (SLPs) who reported greater exposure to EBP in their graduate program and/or CF experience were also more likely to view EBP more positively. Zipoli and Kennedy also found that SLPs who reported reading research articles and conducting database searches during graduate school were more likely to continue to do so in their professional practice. Of the three components of EBP, research evidence, clinical expertise, and client preferences, SLPs reported they relied on their clinical experience and the expertise of colleagues most often. Dollaghan (2004) argued that empirical findings, if available, should be emphasized over expert opinion in terms of evidence quality; however, it is clear that clinicians should also incorporate clinical expertise and client preferences to guide clinical decision making. Finally, Zipoli and Kennedy noted that SLPs most often reported a lack of time as the number one barrier to providing EBP.

Until a recent publication by Thome et al. (2020) , there had been surprisingly few follow-up EBP studies of the same magnitude and scope conducted over the past 15 years since Zipoli and Kennedy published their report. Many of these follow-up studies targeted specific groups of SLPs—for example, those practicing in the schools ( Hoffman et al., 2013 ) or from a particular state (e.g., Guo et al., 2008 ). The purpose of this study is to provide a critical update to Zipoli and Kennedy (2005) by assessing how increased exposure to EBP principles affects reported use of EBP and perceived barriers to providing EBP in clinical decision making.

Earlier studies report mixed findings regarding SLP's confidence with EBP principles. While knowledge and skills related to implementing EBP were not found to be barriers by Zipoli and Kennedy (2005) , other reports convey perceptions of insufficient training and uncertainty about what constitutes EBP. For example, Guo et al. (2008) used a needs assessment to identify how SLPs viewed their EBP training. The SLPs surveyed reported they lacked the appropriate training and resources to interpret the research and determine which practices were evidence based. Guo et al. found that only half of the 84 participants reported knowledge of EBP, yet nearly all participants answered questions probing clinical practice suggesting they implement EBP principles in their clinical decision making. Guo et al. surmised SLPs may not share a common definition of EBP and suggested that graduate training programs should support the development of EBP skills into course work and clinical practicums.

In a large study of 2,762 school-based SLPs, Hoffman et al. (2013) concluded that the amount and type of exposure to EBP principles is also important to consider. Although 75% of the SLPs in their study reported having formal training in EBP, 70% of participants reported a need for even more EBP training. For example, the clinicians in the study reported rarely posing and researching EBP questions (on average, 0–2 a year). Hoffman et al. concluded that SLPs, regardless of years of experience, do not rely on a consistent process to ask and answer clinical questions using research evidence.

Finally, Vallino-Napoli and Reilly (2004) investigated clinicians' definition of what EBP means to their practice. Although nearly all respondents (94%) identified applying research results to clinical practice, only 51% of respondents cited clinical experience/expertise as an additional important element of EBP. Even fewer, 28%, of clinicians recognized client preferences as the final component of EBP. Approximately one quarter of clinicians recognized all three essential components of EBP delivery. This contrasts with Zipoli and Kennedy's finding that SLPs most frequently relied upon clinical expertise to deliver EBP. Recent studies by Fulcher-Rood et al. (2020) and Thome et al. (2020) provided updated findings on clinicians' perspectives on the three EBP components. Participants in the Fulcher-Rood et al. study most often identified research as an essential component of EBP. Only two of their 26 participants (8%) stated all three components of EBP: research, clinical experience, and client preferences. Despite the lack of clear identification of all three components, however, the SLPs from the study self-reported using research, experience, and client needs to determine if they should use a practice clinically. Thome et al. (2020) surveyed 176 SLPs across the United States. Only 14% of SLPs identified the three components of EBP.

Together, these studies suggest that while most SLPs agree that EBP is integral to evaluation and intervention, there is a need for explicit, consistent training of EBP ethos and a process for asking and answering clinical questions.

EBP Resources

In their study of school-based SLPs, Hoffman et al. (2013) categorized EBP resources as tangible versus intangible. Examples of tangible resources, those that can be purchased or printed, include a computer, internet access, journal access, workshops, and continuing education. In contrast, examples of intangible resources, those which cannot be bought or purchased, include dedicated time to research and access to knowledgeable professionals (professional learning communities). We note that, in other EBP studies, dedicated time to research is categorized as a potential barrier to providing EBP ( Fulcher-Rood et al., 2020 ). The majority of SLPs surveyed reported adequate access to tangible resources, such as computers with internet access, but less adequate intangible resources, such as dedicated time for EBP. Additionally, few SLPs reported an established workplace statement regarding EBP practices. Zipoli and Kennedy (2005) also found that EBP and clinical guidelines were not commonly used by practicing clinicians.

Regarding external evidence, Hoffman et al. (2013) reported that, although 87% of respondents were aware that ASHA members have access to all ASHA peer-reviewed journals, they infrequently accessed ASHA journal articles (0–4 times a year) to support EBP. In contrast, Harding et al. (2014) found that “health clinicians,” or clinical practitioners across seven allied health disciplines, accessed databases once a month to research EBP. Several studies have suggested that clinical experience is the most frequently relied upon source of EBP ( McCurtin & Clifford, 2015 ; Zipoli & Kennedy, 2005 ). For example, respondents to McCurtin and Clifford's 2015 survey most often turned to the opinions of colleagues and experts to provide EBP.

A frequently noted barrier to EBP is insufficient time ( Fulcher-Rood et al., 2020 ; Harding et al., 2014 ; Hoffman et al., 2013 ; O'Connor & Pettigrew, 2009 ; Thome et al., 2020 ; Zipoli & Kennedy, 2005 ). Hoffman et al. (2013) reported that less than 10% of their school-based SLP respondents had dedicated time for EBP research and review. Respondents who were afforded dedicated time for research reported, on average, less than 1 hr per week. Considering that posing and answering clinical questions is estimated to take between 3 and 7 hr ( Brackenbury et al., 2008 ), this is clearly insufficient. Other oft reported barriers include high caseload/workload and lack of funds for resources such as journal articles ( Fulcher-Rood et al., 2020 ; Harding et al., 2014 ; Upton & Upton, 2006 )

The literature also reveals barriers related to specific components of EBP training, such as skills to search for and analyze research articles. While some studies probed perceptions of the adequacy of general training of EBP processes ( Harding et al., 2014 ; Hoffman et al., 2013 ), other studies queried whether clinicians possessed specific skills to implement research findings such as the ability to evaluate the results or comprehend statistical analyses. Metcalfe et al. (2001) reported that 78% of clinicians felt ill-equipped to evaluate statistical findings. Similarly, O'Connor and Pettigrew (2009) found 73% of less experienced and 43% of more experienced SLPs had difficulty understanding statistical analyses. Most recently, however, Thome et al. (2020) noted that most SLPs (73%) felt very to somewhat knowledgeable about accessing online databases and 84% were very to somewhat confident in interpreting findings from studies. Their recent finding may reflect EBP training routinely incorporated into graduate school curriculums—a suggestion we will explore in this study.

In 2015, ASHA's EBP committee was replaced with the Committee on Clinical Research, Implementation Science, and Evidence-Based Practice (CRISP) to fulfill a broader aim of upholding principles of EBP and expanding clinical practice research to advance treatment approaches ( ASHA, 2019 ). As EBP principles have become integrated into clinical training and practice over the past decade, there has been a dramatic increase in EBP resources and other tools available to clinicians to help incorporate current research findings into their clinical practice. Given the widely recognized research-to-practice gap that exists in the field, an emerging field of study, implementation science, is dedicated to bridging this gap and enhancing the dissemination of research findings into clinical practice (e.g., Olswang & Prelock, 2015 ). Resources such as ASHA Evidence Maps, websites dedicated to speech-language pathology training and materials, online access to journals, and searchable databases are a few examples of more readily available tools to promote the implementation of practices that are evidence based. Logically, increased exposure to EBP principles in clinical training combined with an increased availability of EBP resources should translate into increased acceptance, understanding, and use of EBP along with fewer perceived barriers to implementation of EBP. However, few studies have been undertaken in recent years to empirically examine this assertion. The purpose of this study is to provide a critical update on current EBP practices in the field of speech-language pathology. We surveyed practicing SLPs to determine how SLPs are using EBP in their clinical practice and their perceptions of barriers to providing EBP to assess the hypotheses that greater exposure to EBP in clinical training coupled with increased availability of EBP resources has led to increased EBP acceptance and use and reduced perceived barriers toward implementing practices that are evidence based.

Participants

We targeted multiple sources including social media, ASHA Special Interest Groups, and state speech and language associations to recruit practicing SLPs from diverse workplace settings to complete the survey. Participants accessed the survey through an electronic link that provided them with a brief description and purpose of the study. Participants were informed that their responses were anonymous, as no data linking a participant to their responses was collected. Informed consent was not explicitly requested because this study qualified as exempt research by the internal review board at Purdue University.

We received 324 responses to the survey; however, we excluded abandoned surveys and responses from participants who had not completed their CF. This resulted in a final data set of 317 participants. Forty-four percent of participants had been employed as an SLP for more than 20 years. Nearly 28% had practiced between 11 and 19 years, and 27.8% had practiced for ≤ 10 years. Most participants, 86.4%, were master's level clinicians, 1.3% of participants held a clinical doctorate and 11.7% of participants held a PhD. Seventy-seven percent of participants indicated they worked full time, and 19.2% indicated they worked part time. Nearly 95% of our participants were ASHA members. Figure 1 illustrates participants' primary practice settings, with the most common setting being elementary, middle, or high schools (37%).

Primary workplace setting for survey respondents. Percentages of SLPs employed by workplace setting. The most common write-in responses for the “Other” category included telepractice, specialty clinics, and home health. Percentages yield greater than 100% accounting for practitioners employed in more than one setting.

We developed a 48-item electronic survey that was modeled after previously published EBP surveys in the field of speech-language pathology to facilitate comparison with earlier findings (e.g., O'Connor & Pettigrew, 2009 ; Vallino-Napoli & Reilly, 2004 ; Zipoli & Kennedy, 2005 ). The first section of the survey gathered demographic information such as participant's age, highest degree, years in practice, employment status, and primary practice setting. The second section probed exposure to EBP during graduate school and CF training. The third section asked about the type and amount of EBP training participants had received during their career. The next section probed EBP use by asking participants to indicate the frequency (i.e., daily, weekly, monthly, yearly, never) that they incorporated different sources of EBP into their clinical practice. These sources represented the three components of EBP—current research findings, clinical experience, and client preferences. The last section probed participants' attitudes and perceived barriers toward EBP by asking their level of agreement on a 5-point scale ranging from strongly disagree to strongly agree .

Six clinical professors of speech-language pathology employed at Purdue University independently reviewed an initial survey draft and provided oral or written feedback. Feedback included suggestions for additional questions and revising question wording to enhance clarity and streamline the survey. The final version of the survey incorporated the changes suggested by this panel of reviewers (see Appendix ).

Predictors of EBP Use

Participants' responses were analyzed using descriptive and inferential statistics. We conducted a standard multiple regression (enter-method) to determine which factors predicted use of EBP. We used the survey items probing sources of EBP in clinical practice in Section 4 along with two items from Section 5, “I routinely incorporate new research findings into my clinical practice” and “I consider my clients' preferences when making clinical decisions” as our planned outcome variable, “EBP use” integrating the three essential components of EBP. The independent variables, years of practice, degree, EBP exposure during graduate program and CF, EBP career training, average barrier score, and average EBP attitude score were used in the regression to predict EBP use. The coefficient of determination, R 2 , represents the proportion of variance in the dependent variable that is explained by variables in the regression model. Standard interpretation of the coefficient of variation— R 2 of .20 = small effect, R 2 of .50 = moderate effect, and R 2 of .80 = large effect—were applied ( Cohen, 1988 ). The independent variable, EBP exposure during graduate training and CF, was calculated by taking the total number of affirmative responses to each item in the second section of the survey with one additional item from the last section probing attitudes/barriers. Specifically, an additional point was added if participants indicated a degree of agreement (i.e., agree or strongly agree) to the statement, “The SLPs that I was in contact with during my CF incorporated EBP into their clinical practice.” Finally, to establish the independent variable EBP career training, the average response was taken across items in the third section of the survey. We assessed the internal consistency of the composite outcome variables (i.e., variables calculated using multiple items from the survey), “EBP use,” average barrier score, and average attitude score using Cronbach's alpha. A reliability coefficient of .70 or higher is considered acceptable internal consistency.

Post Hoc Analysis

Given that time is an often-cited barrier to implementation of EBP, we conducted Pearson correlations to examine the relationship between time and other barriers to assess the hypothesis that individuals citing time as a significant barrier to EBP would also be more impacted by other barriers. We also examined relationships between time and EBP graduate training and workplace training to explore the suggestion that clinicians who received increased EBP training would be less likely to cite time as a significant barrier to implementing EBP in their clinical practice.

The composite outcome variable, “EBP use,” achieved an acceptable internal consistency of 0.79 using Cronbach's alpha. We also examined Cronbach's alpha levels for two of the predictor variables that represented composite scores. The average barrier score achieved an acceptable internal consistency of 0.72. However, the average composite attitude score achieved unacceptably low reliability with coefficients ranging from –.02 to .25 (depending upon whether any of the four items were removed from the overall composite score). Therefore, the predictor, “average attitude score” was removed from the multiple regression model. The remaining five predictors, years of practice, degree, EBP exposure during graduate program and CF, EBP career training, and average barrier score, collectively accounted for a significant percentage, approximately 17%, of the variance of EBP use, F (5, 312) = 12.20; R 2 = .17, p < .001. The coefficient of determination R 2 represents a medium effect ( Cohen, 1988 ). The results of the multiple regression are provided in Table 1 . The correlation matrix for the six predictor variables and EBP use are presented in Table 2 . The three independent variables, exposure to EBP during graduate training and CF, EBP career training, and average barrier score each significantly predicted EBP use.

Multiple regression analysis for predictors of evidence-based practice (EBP) use.

Note.  CF = clinical fellowship.

Correlation matrix ( n = 317).

Note.  EBP = evidence-based practice; CF = clinical fellowship.

We found that an increase in exposure to EBP principles during graduate school/CF and EBP career training each translated into an increase in average EBP use. Table 2 reveals modest correlations between EBP use and these predictors. Of the possible sources of EBP training in graduate school/CF exposure (i.e., dedicated EBP course, individual project related to EBP, EBP training embedded within a course, and EBP training embedded within clinical practicum), most participants indicated that they had received EBP training as part of a class in graduate school (65.4%). Approximately 89% of participants reported at least one affirmative response to EBP training in graduate school/CF. Approximately 87% of participants reported that they had received EBP training during their career (i.e., workplace training, online training, workshop/conference/seminar). The most often cited source of career EBP training was through workshops or conferences, with 95% of respondents selecting this option.

Finally, higher barrier scores were also associated with increased EBP use. It is important to note that higher barrier scores are interpreted positively, meaning the responder was less impacted by a particular barrier (e.g., indicating strong agreement to the statement “I have sufficient access to journal articles at my workplace” would yield a higher score—see Appendix ). Table 2 confirms a moderate correlation between barrier score and EBP use.

Sources of EBP

Participants were asked to indicate the frequency they turned to different sources of EBP with higher scores indicating more frequent use (see Table 3 ). The three overall components of EBP, client preferences, external evidence, and clinical experience represented the top three sources of EBP, respectively. SLPs considered their clients' preferences more frequently than any other source of EBP. Incorporating research into clinical practice earned the second highest score, indicating that SLPs turn to this source on a weekly basis. Relying on clinical expertise received the third highest score. The three most infrequently utilized sources of EBP were posting questions on listservs, presenting findings at conferences/workshops and completing online case studies.

Evidence-based practice (EBP) sources—average scores are based on Likert-scale frequency responses: 0 = never , 1 = yearly , 2 = monthly , 3 = weekly , 4 = daily.

Note.  ASHA = American Speech-Language-Hearing Association.

Barriers to Providing EBP

Table 4 reveals the frequency of participants indicating disagreement (“2”) or strong disagreement (“1”) to each barrier type along with the mean and standard deviation for each barrier score. Recall that disagreement is interpreted negatively. Participants who responded neutrally––neither agree/disagree (“3”) or indicated some level of agreement (“4” or “5”)—were not included in the frequency counts. The two barriers earning an average score of less than 3 were allocated time and caseload size. Over 54% of respondents considered time to be a significant barrier, while over 43% considered caseload size to be a significant barrier. Post hoc correlations indicated a significant relationship between lack of time and several barriers in Table 4 . Participants indicating that time was a significant barrier were more likely to rate caseload size, r (307) = .47, p < .01; access to resources, r (309) = .36, p < .01; lack of training, r (309) = .26, p < .01; and workplace culture, r (309) = .29, p < .01, as barriers to delivering EBP. Although none of the types of EBP training during graduate school correlated with the perception of time as a barrier to implementing EBP (all r –.003 to .08; all p > .05), the barrier time was significantly correlated with EBP workplace training, r (268) = .26, p < .01; online EBP training, r (259) = .18, p < .01; and EBP workshop/conference/seminar training, r (297) = .16, p < .01. Participants who were less likely to rate time as a barrier to providing EBP reported receiving more EBP career training.

Perceived barriers to EBP.

Note.  Items indicating disagreement or strong disagreement (i.e., lower scores on the 5-point scale) were considered significant barriers. Scoring was reversed for negatively stated items (e.g., “I do not have allocated time at work to research/read about my clients”). ASHA = American Speech-Language-Hearing Association.

Over one-third of participants also indicated that access to journal articles at home and comfort with statistical analyses were significant barriers. SLPs infrequently cited the ability to perform a literature search, lack of training, and workplace culture as barriers to providing EBP; these three potential barriers earned the highest scores.

Attitudes Toward EBP

Although we were unable to use the composite score representing average attitude toward EBP in the multiple regression model, we examined the relationship between individual items conveying opinions about EBP use through Pearson correlations and descriptive statistics. Most participants viewed EBP favorably, with 89.3% indicating they somewhat or strongly agreed with the statement, “I am an advocate of EBP.” However, the correlation between average EBP use and advocacy was not significant, r (315) = .05, p = .35. The percentage of participants disagreeing with the statement, “I am uncertain what necessarily constitutes EBP,” reached 79.5%, indicating that most participants had an overall understanding of EBP principles. Greater certainty of what constitutes EBP was associated with increased EBP use, r (312) = .18, p = .001. Overall, most participants, 81.4%, expressed that they were confident in their ability to determine the optimal intervention for their client in the face of conflicting evidence. The correlation between this item relating to confidence and EBP use was significant, r (312) = .12, p = .04. Finally, we noted a range in level of agreement to the statement, “I should increase the use of evidence in my clinical decisions.” Out of 314 responses, 11.7% strongly agreed, 42.3% somewhat agreed, 26.8% were neutral, 12.3% somewhat disagreed, and 6% strongly disagreed. The correlation between this item and average EBP use was not statistically significant, r (312) = –.09, p = .11.

The intent of this study was to document current EBP practices in the field of speech-language pathology by surveying practicing clinicians about their use of EBP in clinical practice, training in EBP, attitudes toward EBP, and perceived barriers to providing EBP. Our survey was modeled after those implemented in previous studies to assess whether increased exposure to EBP principles and access to sources of EBP has led to an increase in EBP use and decrease in perceived barriers. Training and exposure to EBP in graduate school and during the CF, training/exposure in the workplace, and barrier scores each significantly predicted EBP use. Overall, survey respondents reported positive attitudes toward EBP, demonstrated an understanding of EBP, and consistently applied the three components of EBP in their clinical practice.

This study confirms an earlier finding from Zipoli and Kennedy (2005) that exposure and training in EBP during graduate school/CF significantly predicts EBP use. Respondents who indicated more exposure during graduate school/CF were more likely to implement EBP principles in clinical practice. We also queried EBP training in the workplace to clarify how exposure during different career phases supports EBP use. An increase in EBP exposure through workplace training, independent online training, and/or through workshops and conferences also significantly predicted higher EBP use in clinical practice. These findings confirm the importance of EBP exposure during the initial and later stages of clinicians' careers to promote the incorporation of EBP principles into conventional practice.

Finally, barrier scores significantly predicted EBP use. As expected, clinicians indicating that they were less impacted by the practical barriers listed in Table 4 were more likely to report higher use of EBP. It is important to note that, although these three predictor variables accounted for a significant degree of the variability in EBP usage, clearly, there are additional factors we did not explore that contribute substantially to EBP use. We present directions for future research in the following sections to explore additional factors that affect EBP use.

Previous studies revealed a high reliance on clinical expertise; multiple studies found that clinicians relied on their own clinical judgment and the input of qualified colleagues to deliver EBP ( McCurtin & Clifford, 2015 ; Nail-Chiwetalu & Bernstein Ratner, 2007 ; O'Connor & Pettigrew, 2009 ; Togher et al., 2011 ; Zipoli & Kennedy, 2005 ). Although respondents to our survey also cited clinical experience and colleagues' opinions as often turned to sources of EBP, they ranked client preferences and research findings higher, as the No. 1 and 2 sources, respectively. Taken together, our results confirmed frequent––weekly or monthly––application of three principles of EBP, current research findings, clinical experience, and client preferences into practice. This finding may reflect a shift from relying primarily on clinical experience to the inclusion of research findings and client needs/preferences. It is possible that increased training and exposure to all three components of EBP has precipitated this change.

On the other hand, this result seems at odds with earlier findings suggesting that only 14%[en dash]25% of SLPs correctly identified the three components of EBP ( Thome et al. (2020) ; allino-Napoli & Reilly, 2004 ). Yet, Vallino-Napoli and Reilly's study predates a time when principles of EBP were routinely incorporated into graduate education programs. Participants in the recent Thome et al. (2020) studies were given a list of five possible answers that included the three components of EBP along with two foils and were instructed to select all that applied. It is important to note that many of the respondents correctly identified the three EBP components, but also selected the foils, accounting for the much lower reported understanding of EBP. When examining each possible answer individually, 97% identified research, 72% identified clinical expertise, and 57% identified client values. Our results align with studies suggesting that clinicians do indeed incorporate evidence-based principles into their clinical practice even if they cannot explicitly identify EBP components ( Fulcher-Rood et al., 2020 ; Guo et al., 2008 ).

There has been an increase in online EBP resources available to SLPs over the past decade. Interestingly, these comprise the more frequently reported (i.e., journal articles, internet browsing, blogs) and the least frequently reported (i.e., online case studies, listservs/e-mail lists) EBP sources (see Table 3 ). While many online sources are available to support EBP use, clearly, clinicians report a greater reliance on some sources over others. Participants in the study by Thome et al. (2020) ranked ASHA resources as being those most likely to be accessed and most helpful. The 2020 pandemic has brought an abrupt shift to telehealth practices. Many SLP master's students, for example, have had to rely on case simulation and teletherapy as part of their clinical caseloads. This may result in greater reliance on online EBP resources in the coming years.

Current EBP Barriers

Corroborating other studies, insufficient time and workload/caseload size remain the most often cited barriers to EBP implementation, with over half of respondents indicating they had insufficient time to research topics, and over 43% citing workload/caseload size as barriers to providing EBP ( Fulcher-Rood et al., 2020 ; Harding et al., 2014 ; Hoffman et al., 2013 ; O'Connor & Pettigrew, 2009 ; Sadeghi-Bazargani et al., 2014 ; Vallino-Napoli & Reilly, 2004 ; Zipoli & Kennedy, 2005 ).

Harding et al. (2014) espoused the importance of not only identifying barriers to implementing EBP but elucidating the nature of these barriers to effect changes that ultimately lead to an increase in EBP. It was unsurprising to find a positive relationship between time and caseload size as our results revealed. However, it is noteworthy that many clinicians who cited time as a significant barrier also indicated inadequate support in the workplace including lack of training and access to resources and workplace culture as additional hurdles to providing EBP. After an SLP has identified an EBP, the next step is to effectively apply it to their clinical practice. Implementation science has sought to identify strategies that promote the more efficient transfer of research findings into practice (e.g., Olswang & Prelock, 2015 ). Investigations into effective strategies for implementation of research findings often cite the need for support and guidance from employers/organizations ( Douglas et al., 2015 ; Powell et al., 2015 ).

While it could be argued that time to pose and answer clinical questions through research review is a compulsory component of clinical services, current demands on clinicians do not support this ideal. Health clinicians in the Harding et al. (2014) study reported that research and review for EBP was viewed as a part of professional development as opposed to an essential part of the workload, resulting in limited administrative support for research time within the workday. Limited time to ask and answer clinical questions is a direct result of requirements for productivity. Nonbillable indirect client services are often neglected in the establishment of workload responsibilities. In school settings, this translates to time required to meet student needs that cannot be counted as service minutes. Further integration of dedicated time to ask and answer clinical questions into workload is needed to reduce the barrier of time.

Previous studies identified insufficient training as a barrier to providing EBP. For example, Hoffman et al. (2013) found that 25% of school-based SLPs reported no formal EBP training. Other studies cited insufficient training in specific areas such as statistical analysis as a barrier. O'Connor and Pettigrew (2009) noted that a significant proportion, or 47%, of respondents reported difficulty understanding statistical results/analysis. We found that a smaller proportion––16% of respondents—identified a lack of training as a barrier to providing EBP. Fewer SLPs, or 33%, identified difficulties with statistical analysis as a barrier. The lower percentages in our study compared with earlier reports may reflect, in part, intentional training in research methodology in SLP graduate programs and increased exposure to research in the CF and beyond. However, given that one third of SLPs still view statistics as a barrier to EBP indicates a need for even more focused training in statistical analysis to facilitate interpretation of research findings.

Despite the fact that there were many more online resources available to clinicians when our study was conducted, clinicians still report limited access to online sources of EBP, a finding noted in several earlier reports ( O'Conner & Pettigrew, 2009 ; Vallino-Napoli & Reilly, 2004 ; Zipoli & Kennedy, 2005 ). Again, the 2020 pandemic and resultant move to telepractice could have a long-term impact on online resource availability and use. Access to free webinars and other materials was pervasive at the start of the pandemic. For example, in response to the public health crisis, ASHA made online access to continuing education, the ASHA pass, free to all members for 3 months. Access to this pass afforded practicing SLPs the opportunity to take courses promoting evidence-based telepractice they may not have had the opportunity to take otherwise.

We found that most respondents, over 89%, advocated for EBP, corroborating overall positive attitudes toward EBP by other studies (e.g., Alhaidary, 2019 ; Fulcher-Rood et al., 2020 ; Thome et al., 2020 ; Zipoli & Kennedy, 2005 ). Encouragingly, nearly 80% of respondents indicated they understood what constitutes EBP, with over 81% indicating they could select an appropriate treatment when there is conflicting evidence. This demonstrates an overall understanding of EBP principles and degree of confidence in implementing EBP by most respondents.

Although responses to items assessing attitudes toward EBP were generally positive, we noted a range in responses to the item, “I should increase the use of evidence in my clinical decisions.” Interpreting this range in responses is difficult. For example, it is possible that participants expressing neutrality or disagreement may feel that they already incorporate a sufficient degree of evidence into their clinical decisions. On the other hand, some respondents may feel that they are not able to increase their use of EBP in clinical practice given practical barriers such as time.

Considerations and Limitations

We did not ask participants to report their sex, race/ethnicity, or the state that they practice in. The latest ASHA 2019 profile reveals that over 96% of SLPs are female with 92% of members not identifying as a member of a minority group. Given the homogeneous demographics of ASHA-affiliated SLPs of which nearly all survey participants were members, we would have been underpowered to analyze these data in a meaningful way. We acknowledge, however, that we could have collected and reported these data to serve as a basis of comparison for future studies. The recent study by Thome et al. (2020) included these data along with U.S. regions and whether respondents practiced in rural, urban, or suburban settings, although they did not incorporate these demographic details into their hypotheses or statistical analyses.

Our survey was designed to provide an updated broad perspective of EBP rather than an in-depth probe of individual topics related to EBP. For example, although we asked SLPs to indicate whether they understood what constituted EBP, we did not explicitly ask them to define EBP or to list the components of EBP to document level of comprehension. Similarly, we asked clinicians whether items were barriers to providing EBP; however, we did not ask participants to detail why an item was necessarily a barrier. Follow-up research into prevalent barriers is crucial toward the development of viable solutions to lessen their impact and advance implementation theory and practice ( Powell et al., 2017 ). For example, although clinicians recognize the importance of external evidence, how often do they retrieve articles that help develop effective treatment plans? Building a pipeline of research findings to implement into clinical practice requires developing a substantial evidence base, yet a recent study re-affirms the paucity of clinical practice research publications. Roberts et al. (2020) reported that clinical practice research comprises only 25% of the articles published in ASHA journals over the past 10 years—and these articles were not distributed evenly across disorders. A direction for future study is assessing the initiatives that support and the barriers that hinder clinical practice research to narrow the research-to-practice gap.

Implications and Future Directions

There have been relatively few large-scale studies since Zipoli and Kennedy published their seminal study on EBP in speech-language pathology 15 years ago. Current findings suggest that clinicians' understanding and acceptance of EBP principles is shaped, in part, by laying the groundwork in graduate school and during the CF with continued opportunities for training and exposure in the workplace. Greater exposure to EBP in each of these settings was associated with higher reported use of EBP in clinical practice. While earlier studies identified a primary focus on clinical expertise to deliver EBP, we found the three most frequently cited sources of EBP comprised each of the three principal components: considering client preferences, incorporating new research findings into practice, and reliance on clinical expertise. Our field increasingly recognizes the importance of client and caregiver input into the therapy process to cultivate engagement and, ultimately, improved long-term therapy outcomes ( Braden et al., 2018 ; Jahromi & Ahmadian, 2018 ; Lawton et al., 2018 ; McCoy et al., 2019 ). Strides made to increase client input and feedback should be continued and expanded so that therapeutic alliance is an integral component of the initial assessment and plays an essential role in improved long-term outcomes.

Although we documented increased understanding and use of EBP, long-standing barriers remain. The most frequently cited barrier was insufficient time for research. An understanding of the process and time required to pose and answer clinical questions seems essential to developing strategies to reduce its impact as a barrier to EBP. Brackenbury et al. (2008) estimated that it takes 3–7 hr to pose and answer a clinical question. Considering this estimate, the amount of time (if any) that most clinicians are allotted for research is plainly insufficient.

Reiterating the importance of support from employers and organizations toward implementing EBP, we found a significant relationship between time and EBP career training. Clinicians indicating that they were less impacted by the time barrier were more likely to have received more EBP career training. Follow-up research is needed to clarify this relationship by delineating the type of EBP career training that clinicians receive and the potential impact on their professional time. For example, whether SLPs were trained on techniques that are evidenced based or on specific strategies to support implementation of EBP.

Recent evidence suggests that promoting the efficacy of a treatment does not ensure that it will be adopted into routine clinical practice ( Bauer & Kirchner, 2020 ). Douglas et al. (2015) identified effective strategies that may readily lead to changes in implementing practices that are evidence based, including informal knowledge sharing, onsite coaching/mentoring, consistent access to data/feedback of performance, positive reinforcement, and organizational support. Thus, investment in the implementation of EBP from all stakeholders (SLPs, workplace administrators, educational institutions/research centers, and professional organizations) is essential to advance EBP use.

Considering EBPs result in measurable returns, for example, improved health outcomes and a reduction in the cost of care, it is imprudent not to allocate time for research ( Melnyk et al., 2016 ). Resources that are both accessible and time friendly are also necessary to support the consistent use of EBP. The ASHA Evidence Maps ( https://www.asha.org/Evidence-Maps/ ) is one example of a resource created to facilitate EBP by increasing the efficiency of conducting a literature review. Currently, there are 43 topics covered by the Evidence Maps with several additional topics under development. Most participants, approximately 77%, reported using the Evidence Maps. Additional time-saving resources are needed to support implementation of EBP as well as explicit instruction into the process of asking and answering clinical questions ( ASHA, n.d .; Melnyk et al., 2010 ).

Although few respondents, 15.6%, identified lack of training as a barrier to providing EBP, over 33% indicated that they were not comfortable with the statistical analyses in research articles. Graduate school coursework should not be the only opportunity to learn about research methodology. Continuing education opportunities are needed to further the development of clinical research skills after graduate school. Continuing education offerings focusing on statistical approaches or how to appraise research findings, for example, could help clinicians extrapolate findings to their clinical practice. This would help, in part, advance the career-long development and implementation of EBP.

Acknowledgments

We would like to acknowledge support from our colleagues in the Department of Speech Language and Hearing Sciences at Purdue University. Specifically, we thank Teasha McKinley for her patience and attention to detail throughout the survey development and data collection process. We are grateful to the clinical faculty for their valued contributions to the development of the survey.

Section 1 Demographics

What is your age?

○ Under 25

○ 25–35

○ 36–45

○ 46–60

○ Over 60

What is the highest degree you hold in a speech-language or hearing science program?

○ Bachelor's degree

○ Master's degree

○ Clinical Doctorate degree

○ PhD

○ Other –Write In _______________________________

Comments (Optional): _______________________________

How many years have you been in practice since graduation?

○ 0–5

○ 5–10

○ 10–15

○ 15–20

○ 20+

What is your current employment status?

○ Part time

○ Full time

○ Seeking employment in the field

○ Employed outside of the field

○ Currently unemployed

Comments (Optional): ______________________________

Which best describes your primary work setting? Please mark all that apply.

○ Birth to three (0–3) agency

○ School

○ Hospital/Outpatient clinic

○ Hospital/Inpatient

○ Rehabilitation

○ Long-term care facility

○ University/College

○ Private practice

○ Other –Write In ___________________________________

Comments (Optional): __________________________________

Please mark all professional affiliations for which you are currently a member.

○ ASHA noncertified member

○ ASHA Clinical Fellow

○ ASHA CCC member

○ State organization member

○ Student organization

Comments (Optional): _________________________________

Section II. EBP Training Graduate School Training

During your graduate program indicate which of the following you completed. Please mark all that apply.

Section III. EBP Career Training

Select the type and amount of EBP training you have completed during your professional career. Please mark all that apply.

Section IV. Sources of EBP

Please select the amount of time you spend on each activity in your current clinical practice. If an activity does not apply, please skip to the next item.

Daily  Weekly  Monthly  Yearly  Never

Relying on my own clinical experience/expertise

Gaining input from qualified colleagues

Gaining input from experts in the field

Posting questions on listservs or e-mail lists

Attending seminars, conferences, or workshops

Sharing strategies or ideas with my colleagues

Reading textbooks for information

Reading journal articles for information

Reading clinical practice guidelines (e.g., National, state, and/or facility)

Reading professional blogs

Visiting the ASHA Maps website

Visiting other professional websites

Participating in online case studies (e.g., SimuCase or Master Clinician)

Reading special interest group material (e.g., ASHA SIGs)

Reading discipline-wide publications (e.g., ASHA Leader )

Searching research databases (e.g., Cochrane Library)

Internet browsing (e.g., Google or Google Scholar)

Presenting findings from my clinical practice at conferences or meetings.

Section V. Attitudes and Barriers

Please read each statement and indicate your level of agreement.

Strongly disagree Somewhat disagree Neither agree or disagree Somewhat agree Strongly agree

I am an advocate of EBP.

I have sufficient access to journal articles at my workplace.

I have sufficient access to journal articles at my home.

I do not have allocated time at work for research/to read about my clients.

The SLPs that I was in contact with during my Clinical Fellowship incorporated research findings into their clinical practice.

My workload is too large to keep up with latest research.

I am most comfortable relying on my clinical expertise to make clinical decisions.

I consider my clients' preferences when making clinical decisions.

A lack of access to resources hinders my ability to implement EBP.

I am not confident in my ability to appraise research articles.

I am uncertain what necessarily constitutes EBP.

If there is conflicting evidence, I am confident in my ability to determine the optimal intervention for my client.

I am comfortable with most of the statistical analyses in research articles.

A lack of training hinders my ability to implement EBP.

I should increase the use of evidence in my clinical decisions.

The culture at my workplace advocates EBP.

I am not confident in my ability to perform a literature search.

I routinely incorporate new research findings into my clinical practice.

Funding Statement

Funding for this article was provided by the National Institute on Deafness and Other Communication Disorders, grant R01 DC018000, awarded to Sharon L. Christ.

• Alhaidary, A. (2019). Evidence-based practice patterns among speech-language pathologists and audiologists in Saudi Arabia . Communication Disorders Quarterly , 4 ( 4 ), 242–249. https://doi.org/10.1177/1525740119843681 [ Google Scholar ]
• American Speech-Language-Hearing Association. (2004). Report of the Joint Coordinating Committee on evidence-based practice . http://www.asha.org/uploadedFiles/members/ebp/JCCEBPReport04.pdf
• American Speech-Language-Hearing Association. (2019). Committee on Clinical Research, Implementation Science, and Evidence-Based Practice .. https://www.asha.org/About/governance/committees/CommitteeSmartForms/Committee-on-Clinical-Research,-Implementation-Science,-and-Evidence-Based-Practice/
• American Speech-Language-Hearing Association. (n.d.) The Decide Framework . https://www.asha.org/Research/EBP/Evidence-Based-Practice-Toolkit/
• Bauer, M. S. , & Kirchner, J. (2020). Implementation science: What is it and why should I care? Psychiatry Research , 283 , 112376. https://doi.org/10.1016/j.psychres.2019.04.025 [ PubMed ] [ Google Scholar ]
• Brackenbury, T. , Burroughs, E. , & Hewitt, L. E. (2008). A qualitative examination of current guidelines for evidence-based practice in child language intervention . Language, Speech, and Hearing Services in Schools , 39 ( 1 ), 78–88. https://doi.org/10.1044/0161-1461(2008/008) [ PubMed ] [ Google Scholar ]
• Braden, M. N. , van Leer, E. , McConville, K. , & Blakeslee, S. D. (2018). Patient, parent, and speech-language pathologists' perceptions of pediatric voice therapy through interviews . American Journal of Speech-Language Pathology , 27 ( 4 ), 1385–1404. https://doi.org/10.1044/2018_AJSLP-17-0226 [ PubMed ] [ Google Scholar ]
• Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Erlbaum. [ Google Scholar ]
• Council on Academic Accreditation in Audiology and Speech-Language Pathology. (2014). Standards for accreditation of graduate education programs in audiology and speech-language pathology . American Speech-Language-Hearing Association. http://caa.asha.org/wp-content/uploads/Accreditation-Standards-for-Graduate-Programs.pdf [ Google Scholar ]
• Dollaghan, C. A. (2004). Evidence-based practice in communication disorders: What do we know, and when do we know it? Journal of Communication Disorders , 37 ( 5 ), 391–400. https://doi.org/10.1016/j.jcomdis.2004.04.002 [ PubMed ] [ Google Scholar ]
• Douglas, N. F. , Campbell, W. N. , & Hinckley, J. J. (2015). Implementation science: Buzzword or game changer? Journal of Speech, Language, and Hearing Research , 58 ( 6 ), S1827–S1836. https://doi.org/10.1044/2015_JSLHR-L-15-0302 [ PubMed ] [ Google Scholar ]
• Fulcher-Rood, K. , Castilla-Earls, A. , & Higginbotham, J. (2020). What does evidence-based practice mean to you? A follow-up study examining school-based speech-language pathologists' perspectives on evidence-based practice . American Journal of Speech-Language Pathology , 29 ( 2 ), 688–704. https://doi.org/10.1044/2019_AJSLP-19-00171 [ PubMed ] [ Google Scholar ]
• Guo, R. , Bain, B. A. , & Willer, J. (2008). Results of an assessment of information needs among speech-language pathologists and audiologists in Idaho . Journal of the Medical Library Association , 96 ( 2 ), 138. https://doi.org/10.3163/1536-5050.96.2.138 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Harding, K. E. , Porter, J. , Horne-Thompson, A. , Donley, E. , & Taylor, N. F. (2014). Not enough time or a low priority? Barriers to evidence-based practice for allied health clinicians . Journal of Continuing Education in the Health Professions , 34 ( 4 ), 224–231. https://doi.org/10.1002/chp.21255 [ PubMed ] [ Google Scholar ]
• Hoffman, L. M. , Ireland, M. , Hall-Mills, S. , & Flynn, P. (2013). Evidence-based speech-language pathology practices in schools: Findings from a national survey . Language, Speech, and Hearing Services in Schools , 44 ( 3 ), 266–280. https://doi.org/10.1044/0161-1461(2013/12-0041) [ PubMed ] [ Google Scholar ]
• Jahromi, E. M. , & Ahmadian, L. (2018). Evaluating satisfaction of patients with stutter regarding the tele-speech therapy method and infrastructure . International Journal of Medical Informatics , 115 , 128–133. https://doi.org/10.1016/j.ijmedinf.2018.03.004 [ PubMed ] [ Google Scholar ]
• Lawton, M. , Sage, K. , Haddock, G. , Conroy, P. , & Serrant, L. (2018). Speech and language therapists' perspectives of therapeutic alliance construction and maintenance in aphasia rehabilitation post-stroke . International Journal of Language & Communication Disorders , 53 ( 3 ), 550–563. https://doi.org/10.1111/1460-6984.12368 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• McCoy, S. W. , Palisano, R. , Avery, L. , Jeffries, L. , Laforme Fiss, A. , Chiarello, L. , & Hanna, S. (2019). Physical, occupational, and speech therapy for children with cerebral palsy . Developmental Medicine & Child Neurology , 62 ( 1 ), 140–146. https://doi.org/10.1111/dmcn.14325 [ PubMed ] [ Google Scholar ]
• McCurtin, A. , & Clifford, A. M. (2015). What are the primary influences on treatment decisions? How does this reflect on evidence-based practice? Indications from the discipline of speech and language therapy . Journal of Evaluation in Clinical Practice , 21 ( 6 ), 1178–1189. https://doi.org/10.1111/jep.12385 [ PubMed ] [ Google Scholar ]
• Melnyk, B. M. , Fineout-Overholt, E. , Stillwell, S. B. , & Williamson, K. M. (2010). Evidence-based practice: step by step: the seven steps of evidence-based practice . The American Journal of Nursing , 110 ( 1 ), 51–53. https://doi.org/10.1097/01.naj.0000366056.06605.d2 [ PubMed ] [ Google Scholar ]
• Melnyk, B. M. , Ford, L. G. , Thomas, B. K. , Troseth, M. , Wyngarden, K. , & Szalacha, L. (2016). A study of chief nurse executives indicates low prioritization of evidence-based practice and shortcomings in hospital performance metrics across the United States . Worldviews on Evidence-Based Nursing, 2016 , 13 ( 1 ), 6–14. https://doi.org/10.1111/wvn.12133 [ PubMed ] [ Google Scholar ]
• Metcalfe, C. , Lewin, R. , Wisher, S. , Perry, S. , Bannigan, K. , & Moffett, J. K. (2001). Barriers to implementing the evidence base in four NHS therapies: dietitians, occupational therapists, physiotherapists, speech and language therapists . Physiotherapy , 87 ( 8 ), 433–441. https://doi.org/10.1016/s0031-9406(05)65462-4 [ Google Scholar ]
• Nail-Chiwetalu, B. , & Ratner, N. B. (2007). An assessment of the information-seeking abilities and needs of practicing speech-language pathologists . Journal of the Medical Library Association , 95 ( 2 ), 182. https://doi.org/10.3163/1536-5050.95.2.182 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• O'Connor, S. , & Pettigrew, C. M. (2009). The barriers perceived to prevent the successful implementation of evidence-based practice by speech and language therapists . International Journal of Language & Communication Disorders , 44 ( 6 ), 1018–1035. https://doi.org/10.3109/13682820802585967 [ PubMed ] [ Google Scholar ]
• Olswang, L. B. , & Prelock, P. A. (2015). Bridging the gap between research and practice: Implementation science . Journal of Speech, Language, and Hearing Research , 58 ( 6 ), S1818–S1826. https://doi.org/10.1044/2015_JSLHR-L-14-0305 [ PubMed ] [ Google Scholar ]
• Powell, B. J. , Stanick, C. F. , Halko, H. M. , Dorsey, C. N. , Weiner, B. J. , Barwick, M. A. , Damschroder, L. J. , Wensing, M. , Wolfenden, L. , & Lewis, C. C. (2017). Toward criteria for pragmatic measurement in implementation research and practice: A stakeholder-driven approach using concept mapping . Implementation Science , 12 , 118. https://doi.org/10.1186/s13012-017-0649-x [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Powell, B. J. , Waltz, T. J. , Chinman, M. J. , Damschroder, L. J. , Smith, J. L. , Matthieu, M. M. , Proctor, E. K. , & Kirchner, J. E. (2015). A refined compilation of implementation strategies: results from the Expert Recommendations for Implementing Change (ERIC) project . Implementation Science , 10 ( 1 ), 21. https://doi.org/10.1186/s13012-015-0209-1 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Roberts, M. Y. , Sone, B. J. , Zanzinger, K. E. , Bloem, M. E. , Kulba, K. , Schaff, A. , Davis, K. C. , Reisfeld, N. , & Goldstein, H. (2020). Trends in clinical practice research in ASHA journals: 2008–2018 . American Journal of Speech-Language Pathology , 29 ( 3 ), 1629–1639. https://doi.org/10.1044/2020_AJSLP-19-00011 [ PubMed ] [ Google Scholar ]
• Sackett, D. L. , Rosenberg, W. M. , Gray, J. M. , Haynes, R. B. , & Richardson, W. S. (1996). Evidence based medicine: What it is and what it isn't . British Medical Journal , 313 ( 7050 ), 170. https://doi.org/10.1136/bmj.312.7023.71 [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Sadeghi-Bazargani, H. , Tabrizi, J. S. , & Azami-Aghdash, S. (2014). Barriers to evidence-based medicine: A systematic review . Journal of Evaluation in Clinical Practice , 20 ( 6 ), 793–802. https://doi.org/10.1111/jep.12222 [ PubMed ] [ Google Scholar ]
• Thome, E. K. , Loveall, S. J. , & Henderson, D. E. (2020). A survey of speech-language pathologists' understanding and reported use of evidence-based practice . Perspectives of the ASHA Special Interest Groups , 5 ( 4 ), 984–999. https://doi.org/10.1044/2020_PERSP-20-00008 [ Google Scholar ]
• Togher, L. , Yiannoukas, C. , Lincoln, M. , Power, E. , Munro, N. , Mccabe, P. , Ghosh, P. , Worrall, L. , Ward, E. , Ferguson, A. , Harrison, E. , & Douglas, J. (2011). Evidence-based practice in speech-language pathology curricula: A scoping study . International Journal of Speech-Language Pathology , 13 ( 6 ), 459–468. https://doi.org/10.3109/17549507.2011.595825 [ PubMed ] [ Google Scholar ]
• Upton, D. , & Upton, P. (2006). Development of an evidence-based practice questionnaire for nurses . Journal of Advanced Nursing , 53 ( 4 ), 454–458. https://doi.org/10.1111/j.1365-2648.2006.03739.x [ PubMed ] [ Google Scholar ]
• Vallino-Napoli, L. D. , & Reilly, S. (2004). Evidence-based health care: A survey of speech pathology practice . Advances in Speech Language Pathology , 6 ( 2 ), 107–112. https://doi.org/10.1080/14417040410001708530 [ Google Scholar ]
• Zipoli, R. P., Jr , & Kennedy, M. (2005). Evidence-based practice among speech-language pathologists . American Journal of Speech-Language Pathology , 14 ( 3 ), 208–220. https://doi.org/10.1044/1058-0360(2005/021) [ PubMed ] [ Google Scholar ]

There are no new answers.

There are no comments.

• See all questions asked by sabrina102099

• school Campus Bookshelves
• menu_book Bookshelves
• perm_media Learning Objects
• login Login
• how_to_reg Request Instructor Account
• hub Instructor Commons
• Download Page (PDF)
• Download Full Book (PDF)
• Periodic Table
• Physics Constants
• Scientific Calculator
• Reference & Cite
• Tools expand_more
• Readability

selected template will load here

This action is not available.

2.1.4: Components of a Research Project

• Last updated
• Save as PDF
• Page ID 85831

LEARNING OBJECTIVES

• Describe useful strategies to employ when searching for literature.
• Describe why sociologists review prior literature and how they organize their literature reviews.
• Identify the main sections contained in scholarly journal articles.
• Identify and describe the major components researchers need to plan for when designing a research project.

In this section, we’ll examine the most typical components that make up a research project, bringing in a few additional components to those we have already discussed. Keep in mind that our purpose at this stage is simply to provide a general overview of research design. The specifics of each of the following components will vary from project to project. Further, the stage of a project at which each of these components comes into play may vary. In later chapters, we will consider more specifically how these components work differently depending on the research method being employed.

Searching for Literature

Familiarizing yourself with research that has already been conducted on your topic is one of the first stages of conducting a research project and is crucial for coming up with a good research design. But where to start? How to start? In  Chapter 1.3 "Beginning a Research Project" , you learned about some of the most common databases that house information about published sociological research. As you search for literature, you may have to be fairly broad in your search for articles.

I’m guessing you may feel you’ve heard enough about electronic gadget addiction in this chapter, so let’s consider a different example here. On my campus, much to the chagrin of a group of student smokers, smoking was recently banned. These students were so upset by the idea that they would no longer be allowed to smoke on university grounds that they staged several smoke-outs during which they gathered in populated areas around campus and enjoyed a puff or two together.

A student in my research methods class wanted to understand what motivated this group of students to engage in activism centered around what she perceived to be, in this age of smoke-free facilities, a relatively deviant act. Were the protesters otherwise politically active? How much effort and coordination had it taken to organize the smoke-outs? The student researcher began her research by attempting to familiarize herself with the literature on her topic. Yet her search in Sociological Abstracts for “college student activist smoke-outs,” yielded no results. Concluding there was no prior research on her topic, she informed me that she would need an alternative assignment to the  annotated bibliography  I required since there was no literature for her to review. How do you suppose I responded to this news? What went wrong with this student’s search for literature?

In her first attempt, the student had been too narrow in her search for articles. But did that mean she was off the hook for completing the annotated bibliography assignment? Absolutely not. Instead, she went back to Sociological Abstracts and searched again using different combinations of search terms. Rather than searching for “college student activist smoke-outs” she tried, among other sets of terms, “college student activism.” This time her search yielded a great many articles. Of course, they were not focused on prosmoking activist efforts, but they were focused on her population of interest, college students, and on her broad topic of interest, activism. I suggested that reading articles on college student activism might give her some idea about what other researchers have found in terms of what motivates college students to become involved in activist efforts. I also suggested she could play around with her search terms and look for research on activism centered on other sorts of activities that are perceived by some as deviant, such as marijuana use or veganism. In other words, she needed to be broader in her search for articles.

While this student found success by broadening her search for articles, her reading of those articles needed to be narrower than her search. Once she identified a set of articles to review by searching broadly, it was time to remind herself of her specific research focus: college student activist smoke-outs. Keeping in mind her particular research interest while reviewing the literature gave her the chance to think about how the theories and findings covered in prior studies might or might not apply to her particular point of focus. For example, theories on what motivates activists to get involved might tell her something about the likely reasons the students  she  planned to study got involved. At the same time, those theories might not cover all the particulars of student participation in smoke-outs. Thinking about the different theories then gave the student the opportunity to focus her research plans and even to develop a few hypotheses about what she thought she was likely to find.

Reviewing the Literature

Developing an annotated bibliography is often one of the early steps that researchers take as they begin to familiarize themselves with prior research on their topic. A second step involves a literature review in which a researcher positions his or her work within the context of prior scholarly work in the area. A literature review addresses the following matters: What sorts of questions have other scholars asked about this topic? What do we already know about this topic? What questions remain? As the researcher answers these questions, he or she synthesizes what is contained in the literature, possibly organizing prior findings around themes that are relevant to his or her particular research focus.

I once advised an undergraduate student who conducted a research project on speciesism, the belief that some species are superior to or have more value and rights than others. Her research question was “Why and how do humans construct divisions between themselves and animals?” This student organized her review of literature around the two parts of her research question: the why and the how. In the “why” section of her literature review, she described prior research that addressed questions of why humans are sometimes speciesist. She organized subsections around the three most common answers that were presented in the scholarly literature. She used the same structure in the “how” section of her literature review, arranging subsections around the answers posed in previous literature about  how  humans construct divisions between themselves and animals. This organizational scheme helped readers understand what we already know about the topic and what theories we rely on to help make sense of the topic. In addition, by also highlighting what we still don’t know, it helped the student set the stage for her own empirical research on the topic.

The preceding discussion about how to organize a review of scholarly literature assumes that we all know how to read scholarly literature. Yes, yes, I understand that you must know how to read. But reading scholarly articles can be a bit more challenging than reading a textbook. Here are a few pointers about how to do it successfully. First, it is important to understand the various sections that are typically contained in scholarly journals’ reports of empirical research. One of the most important and easiest to spot sections of a journal article is its  abstract , the short paragraph at the beginning of an article that summarizes the author’s research question, methods used to answer the question, and key findings. The abstract may also give you some idea about the theoretical proclivities of the author. As a result, reading the abstract gives you both a framework for understanding the rest of the article and the punch line. It tells you what the author(s) found and whether the article is relevant to your area of inquiry.

After the abstract, most journal articles will contain the following sections (although exact section names are likely to vary): introduction, literature review, methodology, findings, and discussion. Of course, there will also be a list of references cited,Lists of references cited are a useful source for finding additional literature in an area. and there may be a few tables, figures, or appendices at the end of the article as well. While you should get into the habit of familiarizing yourself with articles you wish to cite  in their entirety , there are strategic ways to read journal articles that can make them a little easier to digest. Once you have read the abstract and determined that this is an article you’d like to read in full, read through the discussion section at the end of the article next. Because your own review of literature is likely to emphasize findings from previous literature, you should make sure that you have a clear idea about what those findings are. Reading an article’s discussion section helps you understand what the author views as the study’s major findings and how the author perceives those findings to relate to other research.

As you read through the rest of the article, think about the elements of research design that we have covered in this chapter. What approach does the researcher take? Is the research exploratory, descriptive, or explanatory? Is it inductive or deductive? Idiographic or nomothetic? Qualitative or quantitative? What claims does the author make about causality? What are the author’s units of analysis and observation? Use what you have learned in this chapter about the promise and potential pitfalls associated with each of these research elements to help you responsibly read and understand the articles you review. Future chapters of this text will address other elements of journal articles, including choices about measurement, sampling, and research method. As you learn about these additional items, you will increasingly gain more knowledge that you can apply as you read and critique the scholarly literature in your area of inquiry.

Additional Important Components

Thinking about the overarching goals of your research project and finding and reviewing the existing literature on your topic are two of the initial steps you’ll take when designing a research project. Forming a clear research question, as discussed in  Chapter 1.3 "Beginning a Research Project" , is another crucial step. There are a number of other important research design components you’ll need to consider, and we will discuss those here.

At the same time that you work to identify a clear research question, you will probably also think about the overarching goals of your research project. Will it be exploratory, descriptive, or explanatory? Will your approach be idiographic or nomothetic, inductive or deductive? How you design your project might also be determined in part by whether you aim for your research to have some direct application or if your goal is to contribute more generally to sociological knowledge about your topic. Next, think about what your units of analysis and units of observation will be. These will help you identify the key concepts you will study. Once you have identified those concepts, you’ll need to decide how to define them, and how you’ll  know  that you’re observing them when it comes time to collect your data. Defining your concepts, and knowing them when you see them, has to do with conceptualization and operationalization. Of course, you also need to know what approach you will take to collect your data. Thus identifying your research method is another important part of research design. You also need to think about who your research participants will be and what larger group(s) they may represent. Last, but certainly not least, you should consider any potential ethical concerns that could arise during the course of your research project. These concerns might come up during your data collection, but they might also arise when you get to the point of analyzing or sharing your research results.

Decisions about the various research components do not necessarily occur in sequential order. In fact, you may have to think about potential ethical concerns even before zeroing in on a specific research question. Similarly, the goal of being able to make generalizations about your population of interest could shape the decisions you make about your method of data collection. Putting it all together, the following list shows some of the major components you’ll need to consider as you design your research project:

• Research question
• Literature review
• Research strategy (idiographic or nomothetic, inductive or deductive)
• Research goals (basic or applied)
• Units of analysis and units of observation
• Key concepts (conceptualization and operationalization)
• Method of data collection
• Research participants (sample and population)
• Ethical concerns

KEY TAKEAWAYS

• When identifying and reading relevant literature, be broad in your search  for  articles, but be narrower in your reading  of  articles.
• Writing an annotated bibliography can be a helpful first step to familiarize yourself with prior research in your area of interest.
• Literature reviews summarize and synthesize prior research.
• Literature reviews are typically organized around substantive ideas that are relevant to one’s research question rather than around individual studies or article authors.
• When designing a research project, be sure to think about, plan for, and identify a research question, a review of literature, a research strategy, research goals, units of analysis and units of observation, key concepts, method(s) of data collection, population and sample, and potential ethical concerns.
• Find and read a complete journal article that addresses a topic that is of interest to you (perhaps using Sociological Abstracts, which is introduced in  Chapter 3.1 "Beginning a Research Project" ). In four to eight sentences, summarize the author’s research question, theoretical framing, methods used, and major findings. Reread the article, and see how close you were in reporting these key elements. What did you understand and remember best? What did you leave out? What reading strategies may have helped you better recall relevant details from the article?
• Using the example of students’ electronic gadget addictions, design a hypothetical research project by identifying a plan for each of the nine components of research design that are presented in this section.

• Research Article
• Open access
• Published: 09 April 2024

Constructing eRNA-mediated gene regulatory networks to explore the genetic basis of muscle and fat-relevant traits in pigs

• Chao Wang 1 , 2 , 3   na1 ,
• Choulin Chen 1 , 2 , 3   na1 ,
• Bowen Lei 1 , 2 , 3 ,
• Shenghua Qin 1 , 2 ,
• Yuanyuan Zhang 1 , 2 , 4 , 5 ,
• Kui Li 1 , 2 ,
• Song Zhang 1 , 2 &
• Yuwen Liu 1 , 2 , 3 , 6  

Genetics Selection Evolution volume  56 , Article number:  28 ( 2024 ) Cite this article

407 Accesses

Metrics details

Enhancer RNAs (eRNAs) play a crucial role in transcriptional regulation. While significant progress has been made in understanding epigenetic regulation mediated by eRNAs, research on the construction of eRNA-mediated gene regulatory networks (eGRN) and the identification of critical network components that influence complex traits is lacking.

Here, employing the pig as a model, we conducted a comprehensive study using H3K27ac histone ChIP-seq and RNA-seq data to construct eRNA expression profiles from multiple tissues of two distinct pig breeds, namely Enshi Black (ES) and Duroc. In addition to revealing the regulatory landscape of eRNAs at the tissue level, we developed an innovative network construction and refinement method by integrating RNA-seq, ChIP-seq, genome-wide association study (GWAS) signals and enhancer-modulating effects of single nucleotide polymorphisms (SNPs) measured by self-transcribing active regulatory region sequencing (STARR-seq) experiments. Using this approach, we unraveled eGRN that significantly influence the growth and development of muscle and fat tissues, and identified several novel genes that affect adipocyte differentiation in a cell line model.

Conclusions

Our work not only provides novel insights into the genetic basis of economic pig traits, but also offers a generalizable approach to elucidate the eRNA-mediated transcriptional regulation underlying a wide spectrum of complex traits for diverse organisms.

Significant genetic and phenotypic diversity exists among breeds of domestic animals, which provide ideal models to understand the genetic basis of complex traits. Linking genetic variants to phenotypes using such models is critical in improving livestock production capacity and in elucidating the genetic basis of traits shared by humans and animals. In humans and animals, extensive efforts in mapping causal genetic variants that underlie traits suggest that more than 90% of the single nucleotide polymorphisms (SNPs) that are associated with various complex traits or diseases are located in the non-coding part of the genome [ 1 , 2 , 3 ]. Compared to the biological interpretation of coding variants, our lack of understanding of the genetic code of non-coding regulatory variants has motivated us to initiate several ambitious projects with the aim of systematically characterizing how the non-coding genome functions in various organisms. These projects include ModENCODE for model organisms [ 4 ], ENCODE for humans [ 5 ], and FAANG for farm animals [ 6 ]. These collaborative projects have not only generated multidimensional and temporo-spatial regulatory landscapes in different species, but have also played an important role in identifying the causal non-coding DNA variants and their target genes that would otherwise have been overlooked [ 7 , 8 , 9 , 10 , 11 ].

In spite of the extensive body of research in functional genomics conducted in humans and animals, there remains untapped potential in the comprehensive exploration of the eRNA landscape and the identification of crucial transcriptional regulatory networks mediated by eRNAs. These untapped resources hold great promise for advancing our understanding of the genetic basis of complex traits. Discovered in the 1980s, eRNAs are transcribed from enhancer regions [ 12 ], and their expression levels have been found to be strongly correlated with enhancer activity [ 13 , 14 ]. In vivo, eRNA has multiple regulatory functions, such as mediating the formation of enhancer-promoter loops, maintaining an open chromatin state, changing the chromatin spatial conformation, and recruiting RNAP2 and other transcription initiation complex molecules [ 13 , 15 , 16 , 17 ]. In recent years, the important biological functions of eRNAs have just began to be unraveled in the regulation of diseases, such as cancer, neurodegenerative diseases, cardiovascular diseases, and metabolic diseases [ 18 ]. Therefore, comprehensively analyzing the role of eRNAs, as a hub in integrating signaling from upstream transcription factors (TF) to regulate downstream target genes, would significantly contribute to understanding the genetic basis of complex traits.

The identification of eRNAs presents a significant challenge due to their inherently short length, infrequent splicing, and instability. In spite of these difficulties, the progress of sequencing technologies has provided several methods, such as global run-on sequencing (GRO-seq) [ 19 ], precision run-on sequencing (PRO-seq) [ 20 ], native elongating transcript sequencing (NET-Seq) [ 21 ], cap analysis gene expression and deep sequencing (CAGE-seq) [ 22 ], to identify eRNAs. Although efficient, the complex and costly nature of these methods remains a hindrance to their widespread use. An alternative approach is to use RNA-seq data to identify eRNAs [ 23 , 24 , 25 ]. There has been a growing trend of using this approach to uncover eRNAs in various biological contexts [ 26 , 27 , 28 , 29 , 30 ], including the first animal eRNA database with eRNA annotation in 10 species [ 31 ]. However, the eRNA annotation in pigs has yet to be established.

Among domesticated animals, the pig ( Sus scrofa ) holds a pivotal role, serving not only as a crucial source of protein and fat but also as an exceptional model for biomedical research. More importantly, geographical divergence, local adaptation, and artificial selection have resulted in significant phenotypic differences between eastern (Asia) and western (Europe and America) pigs, including lean meat mass and fat deposition [ 11 , 32 , 33 ]. In the pig husbandry industry, the pursuit of lean meat mass and controlled fat deposition stands as crucial breeding objectives, given their direct impact on meat quality and production efficiency. The Enshi Black (ES) pig, a representative of eastern pig breeds has a stronger fat deposition ability than the western breed Duroc pig. By comparison, the later has a higher lean meat ratio than the former. These contrasting phenotypes offer a unique perspective to study the genetic mechanisms that underlie the development and homeostasis of muscle and fat tissues.

In this study, our aim was to profile the landscape of eRNAs in the muscle and fat tissues of Duroc and ES pigs, respectively. We present potential hub eRNAs and their target genes by an integrative approach combining eRNA-mediated transcriptional regulatory networks, genome-wide association study (GWAS) signal and STARR-seq experiments. STARR-seq is a sequencing-based high-throughput method that allows a direct measure of the allelic regulatory activity of SNPs [ 34 , 35 ]. Our study not only sheds light on the genetic regulatory mechanisms underpinning lean meat mass and fat deposition in pigs, but also proposes an integrative framework to pinpoint eRNA-mediated genetic regulatory networks that might substantially influence complex traits in various species.

Data source

In this study, we used H3K27ac ChIP-seq data obtained from the longissimus dorsi skeletal muscle and subcutaneous fat tissues of 2-week-old Duroc and ES pigs to identify enhancers, with two biological replicates for each breed. In addition, we integrated RNA-seq data from these pigs, covering various tissues such as skeletal muscle ( longissimus dorsi ), subcutaneous fat, heart, liver, and spleen, for a comprehensive quantitative analysis of gene and RNA expression. In addition, we included topologically associated domain (TAD) information for the longissimus dorsi skeletal muscle of two-week-old Large White pigs. All datasets were sourced from the research conducted in Professor Shuhong Zhao’s laboratory under the accession number PRJNA597497 [ 36 ]. To complement our investigation, we accessed GWAS hits and quantitative trait loci (QTL) regions associated with different traits in pigs from pigQTLdb, available at https://www.animalgenome.org/cgi-bin/QTLdb/SS/index . Furthermore, we collected GWAS summary statistics for 64 human complex traits from Hook’s study [ 37 ], accessible at https://doi.org/ https://doi.org/10.5281/zenodo.3253180 . These datasets were selected based on their relevance to our research questions and their availability.

Identification of enhancers and super enhancers

To obtain annotation information on genomic enhancers in the muscle and adipose tissue of Duroc and ES pigs, we acquired H3K27ac signal peaks for identifying the location of enhancers. Each tissue consisted of two biological replicates. In order to ensure consistency in the H3K27ac signal peaks within each tissue, we merged the peaks from the biological replicates using the ‘intersect’ and ‘merge’ commands in BEDTools (version 2.31.0) with the default parameters [ 38 ]. Subsequently, we excluded H3K27ac signal peaks that overlapped with the transcription start site (TSS) of known genes within ± 1 kb and considered the remaining H3K27ac peaks as enhancers. To measure the activity of enhancers in tissues, we downloaded the raw ChIP-seq data of the H3K27ac histone modification for evaluating enhancer activity. First, we processed the ChIP-seq data using the TrimGalore software (version 0.6.7, Babraham Institute, Cambridge, UK) to eliminate sequencing adapters, low-quality bases (Phred < 20), and short sequences (-q 20 --phred33 --stringency 4 --length 25 -e 0.1). After quality control, we aligned the clean data to the susScr11 genome using the Bowtie2 (version 2.4.4) software with the following parameters: --very-sensitive -X 1500 -x genome index -1 fq1 -2 fq2 [ 39 ]. PCR duplicates were removed using the Sambamba (version 1.0.0) [ 40 ] software. From each sample, we extracted pairs of aligned concordant reads, resulting in bam files for quantifying enhancer activity. Referring to Zhao's approach [ 36 ], we used the multiBamSummary BED-file function in deepTools [ 41 ] (v2.088) to count the number of reads within the ± 1 kb region around the center of different tissue enhancers. We then normalized the enhancer read counts by dividing them by the total number of reads in the library. Finally, we assessed the intensity of enhancer activity by calculating the fold-change value (IPRPM /INPUTRPM, where IPRPM represents the normalized signal strength of enhancers in the IP library, and INPUTRPM corresponds to the normalized signal strength of enhancer regions in the INPUT library). The dynamic activity heatmap depicting enhancer activity in fat and muscle tissue was created using the ‘Heatmap’ function from the ‘ComplexHeatmap’ R package (version 2.6.2) within the R software (version 4.0.5).

To identify super-enhancers (SE) in Duroc muscle, Duroc fat, ES muscle, and ES fat tissues, the bam files from biological replicates of each tissue were combined. Next, we used the MACS2 (version 2.2.8) [ 42 ] software to identify H3K27ac peaks specific to each tissue. Subsequently, with the resulting files and the genomic annotation file for the pig, we employed the ROSE (version 0.1) [ 43 ] software using default parameters to pinpoint the SE for each tissue.

Identification of eRNAs

As the transcriptional range of eRNAs can be wider than the enhancer region [ 44 ], we expanded our enhancer set by ± 3 kb around the central point of each enhancer, delineating it as the potential transcriptional activity region of enhancers, in alignment with established methodologies from previous research [ 27 , 29 ]. In order to mitigate interference from known coding genes or non-coding RNAs (such as miRNA, misc_RNA, rRNA, snoRNA, snRNA, tRNA, and lncRNA) during eRNA quantification, we followed the eRNA identification method outlined in Carullo et al. [ 30 ]. In the subsequent transcription signal quantification analysis, our exclusive focus was on the potential transcription regions of enhancers falling more than 1 kb outside of the genes and non-coding RNAs curated by RefSeq, UCSC, and Ensembl databases. Subsequently, we used strand-specific RNA-seq data from two biological replicates of muscle and adipose tissue in ES and Duroc pigs for eRNA analysis. The individuals contributing to the RNA-seq data were the same as those from which the H3K27ac ChIP-seq data were derived. Initially, quality control for the RNA-seq data was conducted using the TrimGalore software (version 0.6.7). The clean data were then mapped to the reference genome (susScr11) using the Hisat2 (version 2.2.1) software [ 45 ], using the (--rna-strandness RF) option for strand-specific mapping. Following the mapping step, we employed the Seqmonk software (Babraham Institute) in accordance with a previously described methodology [ 30 ] to assess the expression levels of eRNAs. In our study, we defined a mean RPM value of ≥ 1 in biological replicates specific to each tissue as the threshold for detectable eRNA expression in that tissue. In addition, we considered the directionality of eRNA transcription. If the proportion of reads mapped to the positive strand of the genome fell between 5 and 95% of the total reads mapped to the enhancer region, the eRNA was classified as bidirectionally transcribed and otherwise, as unidirectional transcribed. To visualize the distribution of detectable eRNAs across various tissues, we used the UpSetR (version 1.4.0) [ 46 ] package in R. In addition, we generated a heatmap of sample-to-sample distances using the expression values of detectable eRNAs, and the pheatmap() function from the pheatmap package (version 1.0.10).

Analysis of eRNA characteristics

To compare the GC content of eRNAs with different transcriptional directions, we used the “nuc” command from the BEDTools software (v2.31.0) to calculate the GC content ratio within each eRNA sequence. Subsequently, we performed a Student's t-test to assess the differences in GC content between the groups of bidirectional and unidirectional eRNAs. The differences in expression level between unidirectional and bidirectional eRNAs, as well as between eRNAs within and outside SE, were evaluated using a two-sided unpaired Wilcoxon test. To explore the relationship between eRNA expression level and enhancer activity in each tissue, the detectable eRNAs were sorted from highest to lowest expression level and divided into eight bins. Then, we used the cor.test function in the R program (version 4.0.5) to assess the correlation between the mean eRNA expression levels and the mean enhancer activity across these bins (method = “pearson”). All the statistical analyses mentioned above were conducted using the R programming language. In our study, we conducted all gene ontology (GO) enrichment analyses using the enrichGO function from the R package clusterProfiler (version 3.14.3) [ 47 ]. Functional enrichment analysis of tissue-specific expression genes and eRNA target genes was directly performed using enrichGO. For eRNA-related functional annotation analysis, we extracted neighboring genes and performed the analysis.

Deep learning model for prediction of enhancers with different transcription patterns

To analyze the sequence features of enhancers with different transcription directions, we used two prominent deep learning classification models: DeepSEA [ 48 ] and a convolutional long short-term memory (LSTM) [ 49 ], to predict the transcriptional direction of enhancers, both being recognized for their efficacy in enhancer sequence prediction. For details on the code and usage of these two deep learning models, please refer to https://github.com/minxueric/ismb2017_lstm . In our analysis, we used the 2-kb enhancer region corresponding to the eRNA as the input sequence for the model. We treated 4832 unidirectional transcribed enhancers as negative samples and 12,883 bidirectional transcribed enhancers as positive samples. To split the dataset, we used 85% of the data for training, 5% for model selection as a validation set, and the remaining 10% for testing the model. We evaluated the performance of the two models on the test data by calculating the area under a receiver operating characteristic (ROC) curve (AUC) value, the ROC curves were plotted using the plotROC [ 50 ] package (version 1.3) in R.

Transposon analysis

To annotate transposons in the pig genome, we used the RepeatMasker [ 51 , 52 ] software (version 4.1.2-p1) and the Repbase-20181026 library (RepeatMasker -parallel 15 -species pig -html -e rmblast -s -a -gff -dir pig_repeat susScr11.fa). We categorized enhancers into two groups: transcribed enhancers (TEn and non-transcribed enhancers (non-TEn). Specifically, enhancers that intersect with the center points of eRNAs were referred to as TEn, while enhancers without intersection with the center points of eRNAs were termed non-TEn. We then performed a comparative analysis of transposon insertions within TEn and non-TEn using the "intersect" command of the BEDTools (v2.31.0) software. First, we examined the differences in transposon insertion frequencies between TEn and non-TEn using a two-tailed Fisher’s exact test. Furthermore, we calculated the base composition of different classes and families of transposons within TEn and non-TEn using R, and all results were visualized using the ggplot2 package. To assess the differential enrichment of transposon families in TEn and non-TEn, we performed a permutation test by simulating elements 1000 times within the genomic background. The resulting p-values were corrected by the false discovery rate (FDR) method. The enrichment results were then visualized using the pheatmap package (version 1.0.10) in R.

Identification of tissue-specific eRNAs and genes

To identify tissue-specific eRNAs and genes, we retrieved and analyzed strand-specific RNA-seq data from the heart, liver, and spleen tissues of both Duroc and ES pigs. It is essential to emphasize that these pigs were the same individuals than those used for enhancer identification in the current study. Each tissue type was represented by two biological replicates. The quality control and alignment methods for the transcriptome data were consistent with those mentioned earlier. Gene counts were quantified using FeatureCounts (version 2.0.2) [ 53 ] on the bam files of each sample. We calculated the gene expression level (TPM) based on the raw count expression matrix, taking gene length and library depth into account using an R script. In addition, we evaluated the expression levels (RPM) of detectable eRNAs in each tissue using the Seqmonk software (Babraham Institute), which is consistent with the previously described quantification methods. Finally, we identified tissue-specific eRNAs and genes for each breed by analyzing the expression profiles across five tissues: muscle, fat, heart, liver, and spleen. Specifically, it is important to note that the tissue specificity of eRNAs was calculated separately for biological replicates 1 and 2. This analysis was conducted using the tissue specificity index (TSI) [ 54 ], which is calculated as follows:

where \({\text{N}}\) is the number of tissues and \({x}_{{\text{i}}}\) is the expression of the eRNA or gene \(x\) in tissue \({\text{i}}\) . Genes or eRNAs with a TSI greater than 0.8 in both biological replicates of the specific tissue in the same breed were deemed to be tissue-specifically expressed in that particular breed.

Enrichment analysis of tissue-specific eRNAs

In this study, we conducted a hypergeometric test to assess the enrichment of tissue-specific expressed eRNAs within super-enhancers in eastern and western pig muscle and adipose tissues. This analysis used key parameters including the total number of detectable eRNAs in each tissue (T), the number of detectable eRNAs within super-enhancers (M), the number of tissue-specific detectable eRNAs (t), and the number of tissue-specific detectable eRNAs within super-enhancers (m). In addition, we performed a hypergeometric test to assess the enrichment of tissue-specific expressed genes within ± 1 Mb regions of tissue-specific eRNAs. In this analysis, X denotes the total number of detectable genes in the tissue, Y represents the count of detectable genes within ± 1 Mb of eRNAs, x indicates the number of tissue-specific expressed genes, and y represents the count of tissue-specific expressed genes within ± 1 Mb of eRNAs.

To explore the potential involvement of TF in regulating tissue-specific eRNAs, we performed motif enrichment analysis using the HOMER [ 55 ] software on H3K27ac histone peaks within the regions of tissue-specific eRNAs. We focused on the top 20 enriched TF motifs within tissue-specific eRNAs and retained only the TF that showed significant enrichment (Q-value < 0.01) and were expressed in the corresponding tissue. To visualize the enrichment levels, we used the ggplot2 package for data visualization.

To investigate the contribution of tissue-specific eRNAs to the heritability of tissue-related biological traits, we obtained datasets from pigQTLdb, which contained GWAS hits and QTL regions associated with various pig traits. These traits were classified into five major classes: Meat and Carcass, Health, Production, Reproduction, and Exterior. After downloading the datasets, we processed and extracted the location information for GWAS hits and QTL regions associated with each trait class. Next, we performed permutation tests by simulating tissue-specific eRNA elements within a genomic background, repeating the process 1000 times (see Additional file 1 Supplementary code). This allowed us to assess the enrichment of GWAS hits and QTL regions associated with different trait classes in tissue-specific eRNAs. It is important to note that we expanded the GWAS hits by ± 20 kb to account for the potential influence of linkage disequilibrium (LD) among SNPs.

To investigate the contribution of pig tissue-specific eRNAs in elucidating the genetic heritability of human diseases, we obtained the susScr11ToHg19.over.chain file from the UCSC database and employed the liftOver tool ( https://hgdownload.soe.ucsc.edu/admin/exe/linux.x86_64/liftOver ) [ 56 ] to convert pig tissue-specific eRNAs into homologous sequence regions in humans. specifying a length of 6 kb and a minimum match threshold of 0.5. Subsequently, we downloaded GWAS summary statistics for 64 traits from Hook’s study [ 37 ]. Using the LDSC method [ 57 ], we dissected the heritability of these 64 complex traits in humans based on the pig-driven human tissue-specific expression of eRNAs. Finally, we combined and analyzed the partitioned heritability contributions for all traits and visualized the results in R.

Construction of an eRNA regulatory network

To construct directed eGRN based on tissue-specific eRNAs, we used a unique approach that involved identifying the upstream TF and downstream target genes of eRNAs. For the identification of upstream TF, we used a combination of eRNA sequence motif scanning and co-expression analysis between TF and eRNA expression levels. Initially, we obtained the custom motif matrices files for TF, which contain motif sequences ( http://homer.ucsd.edu/homer/custom.motifs ). We then extracted the DNA sequence corresponding to the enhancer region of each eRNA from the susScr11.fa file using the “getfasta” command provided by BEDTools (v2.31.0). Motif scanning was conducted using the scanMotifGenomeWide.pl command (scanMotifGenomeWide.pl custom.motifs element.fa -bed) within the HOMER software. TF with motif sequences that could be detected on the DNA sequence and were expressed in the tissue were considered as potential upstream regulators of eRNAs. The confirmation of these potential upstream TF as true regulators of the eRNAs was based on their significant correlation with the regulation of eRNA expression in 20 samples (Rs > 0.5, FDR < 0.05). The correlation analysis was performed using the ‘spearman’ method from the psych package (version 2.2.5) in R, with adjustment for multiple testing using the FDR method. For the identification of target genes of eRNAs, we employed a similar strategy of co-expression analysis between eRNAs and genes in 20 samples. The expression correlation calculation method was consistent with the previous approach, with a threshold set at Rs > 0.3 and FDR < 0.05. In addition, the target genes had to be located within a 1 Mb range of the eRNA. To further filter the target genes of eRNAs, we leveraged the conserved nature of the topologically associated domain (TAD) structure in the 3D genome across species [ 58 ]. We used TAD regions derived from 2-week-old large white pig muscle tissue and required that both the eRNA and its target genes be located within the same TAD region. Lastly, to enhance the elucidation of the regulatory mechanisms that underlie the traits, we integrated population-based GWAS data into our analysis. Specifically, we extracted GWAS hits associated with fat and muscle traits from pigQTLdb (see Additional file 2 : Table S1). To account for the LD effect of SNPs, we specifically screened for eRNAs that intersected with the GWAS hits within a ± 20 kb region. This approach allowed us to identify eRNAs with potential regulatory effects on the corresponding traits. Subsequently, we constructed separate eGRN using the eRNAs associated with muscle and fat traits to uncover the genetic regulatory mechanisms underlying these traits. The resulting eGRN were visualized using the Cytoscape [ 59 ] software. To further refine the eGRN in fat tissue, we performed capture STARR-Seq experiments in mouse 3T3 cells, in order to identify eRNA SNPs that affect enhancer activity. In the end, we removed eRNAs without enhancer-modulating SNPs in the final eGRN of the fat tissue.

Cell culture

Mouse 3T3-L1 fibroblasts (ATCC) were initially cultured in DMEM supplemented with 10% calf serum (B7446, Sigma-Aldrich) to promote cell proliferation. During the induction of cell differentiation, a transition from calf serum to 10% fetal bovine serum (FBS) (10091148, Gibco) was carried out upon the initial phase of cell contact inhibition. The aim of this substitution was to enhance the initiation of lipid droplet formation effectively. Two days after replacement of the FBS, the induction of differentiation was accomplished by incubating the cells in differentiation medium A (DMEM containing 0.5 mM IBMX (I5879, Sigma), 1 μM DEX (D1756, Sigma-Aldrich), and 10 μg/mL insulin (HY-P0035, MedChemExpress), and 10% FBS) for 2 days, followed by differentiation medium B (DMEM containing 10 μg/mL insulin and 10% FBS) for subsequent days with a medium change performed every two days to achieve full differentiation into mature adipocytes (Day 8).

SNP screening for STARR-seq

To identify regulatory SNPs in the fat-related eGRN that regulate differential fat deposition abilities between eastern and western pigs, we established screening criteria for SNPs that were assessed using STARR-seq for regulatory activity. Initially, we obtained genome-wide SNP data from the PigVar database [ 60 ] for eastern and western pig populations. We then overlapped this data with eRNAs to identify the SNP locations within eRNAs. To ensure the detection of SNPs within eRNAs in both eastern and western pig populations collected by our research group, and to assess their potential influence on the differential fat phenotypes between the two groups, we focused on SNPs that showed allele frequency differences between the eastern and western pig populations collected by our team, with a minor allele frequency (MAF) difference exceeding 0.3, and ensuring that the MAF in the whole genome mixed pool of eight eastern and eight western pigs was higher than 0.05. SNPs that met the selection criteria were chosen as candidate SNPs for the STARR-seq assay, allowing for the detection of their regulatory activity on DNA elements. The protocol for the study involving the 16 pigs obtained approval from Huazhong Agricultural University (Protocol code: SYXK(e)2020-0084) and the Institutional Animal Care and Use Committee.

Identification of regulatory SNPs by STARR-Seq

Candidate SNPs with regulatory functions were identified using a previously reported STARR-Seq strategy [ 35 , 61 ]. Initially, custom primers for the 22 enhancer elements containing candidate SNPs were designed and synthesized by Sangon Biotech Co., Ltd (Shanghai, China) (see Additional file 3 : Table S2). To introduce SNP polymorphisms into the screening process, DNA samples extracted from various tissues of eight eastern and eight western pigs were pooled in equal amounts and used as PCR templates. The PCR amplification parameters were 25 cycles consisting of 94 °C for 5 min, 95 °C for 30 s, 55 °C for 30 s, 72 °C for 1 min, and a final extension at 72 °C for 7 min. The reaction mixture, with a total volume of 50 μL, contained 1 μg of substrate and 0.4 μM of primers. The amplified products were purified and sonicated to achieve sizes ranging from 300 to 400 bp. These sonicated fragments were ligated into the hSTARR-seq_ORI vector (#99296, addgene). The recombinant vectors were transformed into E. coli DH10B cells (Life Technologies, Eragny, France) by a Gene Pulser Xcell Electroporation System (Bio-Rad Laboratories, Richmond, CA, USA) to produce the input plasmid library, which then was transfected into 3T3-L1 cells by lipofectamine (jetPRIME, Polyplus 101000046). The STARR-Seq input NGS library was derived from the input plasmid library, whereas the output NGS library was generated from transcripts produced by the input plasmids. To be clear, the input NGS library was created by directly amplifying SNP-containing inserts from the plasmid DNA used for cell transfection, thereby serving as a reference for the initial representation of SNP alleles in the starting plasmid pool. In contrast, the output NGS library assesses the abundance of self-transcribed mRNA originating from insert fragments within the transfected plasmid pool. The experimental details of STARR-Seq are described in previous publications [ 35 , 62 ]. Briefly, to prepare the input NGS library, the input library was prepared using 200 ng of plasmid template, divided into four reactions of 50 μL each. PCR amplification was carried out with the specified primers containing Illumina adaptors [forward: 5ʹ-AATGATACGGCGACCACCGAGATCTACAC-index-ACACTCTTTCCCTACACGACGCTCTTCCGATCT-3ʹ; reverse: 5ʹ-CAAGCAGAAGACGGCATACGAGAT-index-GTGACTGGAGTTCAGACGTG-3ʹ] and the following conditions: 95 °C for 3 min, 98 °C for 20 s, 65 °C for 15 s, 72 °C for 30 s (go to step 2 for 15 cycles, 72 °C for 2 min). To prepare the output NGS library, the transfected cells were harvested after 24 h. Total RNAs were extracted, and mRNAs were enriched by oligodT and reverse transcribed using a plasmid specific primer [CAAACTCATCAATGTATCTTATCATG]. Following reverse transcription, the cDNA products were purified and amplified with primers containing Illumina adaptors to construct the output NGS libraries [forward: 5ʹ-AATGATACGGCGACCACCGAGATCTACAC-index-ACACTCTTTCCCTACACGACGCTCTTCCGATCT-3ʹ; reverse: 5ʹ-CAAGCAGAAGACGGCATACGAGAT-index-GTGACTGGAGTTCAGACGTG-3ʹ]. The PCR amplification was performed using the KAPA enzyme (KK2602, Roche) with the following cycling conditions: 95 °C for 3 min, 98 °C for 20 s, 65 °C for 15 s, 72 °C for 30 s (go to step 2 for 18 cycles, 72 °C for 2 min). Both the input NGS library and the output NGS libraries were sequenced using the PE150 strategy on the NovaSeq 6000 platform. In this study, we obtained four output NGS libraries (the RNAs from each biological replicate generated two output libraries) and one input NGS library, the output libraries yielded a range of 2.8 to 3.4 million reads each, with the input library containing 8.5 million reads. After performing quality control and removing PCR duplicates, the reads from the two biological replicates were highly correlated (see Additional file 4 : Figure S1a), Therefore, we merged the two replicates for further analysis to increase statistical power for calling regulatory SNPs. When calculating the effect size of SNPs, only SNPs with a coverage exceeding 20 in both input and output libraries were considered. The size of the effect of each SNP was determined by calculating the fold change in allele ratios (output/input). Significance of the differences in allele ratios was assessed using a two-tailed Fisher's exact test. The raw sequence data of STARR-seq used in this paper have been submitted to the Genome Sequence Archive (GSA; https://ngdc.cncb.ac.cn/ ) under accession number CRA011292.

Gene function validation by siRNA

The siRNAs targeting the RB1 , SLC27A6 , RGMA , APLF , UBTD2 genes, and the negative control were custom-synthesized by Ribo Biological Co., Ltd (Guangzhou, China) (see Additional file 5 : Table S3). 3T3-L1 cells were seeded in a six-well culture plate and transfected at 80% confluency. The medium was refreshed 4 to 6 h after the transfection process. To ensure continuous suppression of the target gene expression at a low level, a subsequent knockdown was conducted on the fourth day of the induced differentiation process, with a repeat transfection of siRNA. Then, RNA was extracted using the TriZol reagent, purified by chloroform extraction, precipitated with isopropanol, and dissolved in RNase-free water. Subsequently, 1 μg of RNA was used for cDNA synthesis through reverse transcription in a 20-μL reaction. ChamQ Universal SYBR qPCR Master Mix (Q711 Vazyme) was used for RT-qPCR, 1 μL of cDNA as templates and 0.2 μM primers. The PCR conditions included an initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 10 s, 60 °C for 30 s, and a final step of 95 °C for 15 s, 60 °C for 60 s, and 95 °C for 15 s. The degree of differentiation was assessed by fluorescence microscopy and reverse transcription (RT)-qPCR analyses. Finally, we used the GraphPad Prism 5 software to perform a t-test to compare the expression levels of target genes and marker genes with those of the negative control, and then visualized the expression levels accordingly.

Oil red O staining

When 3T3-L1 cells were fully differentiated into adipocytes (after 8 days), cells were fixed with a 4% formalin solution (BL539A, biosharp) for 15 min at room temperature. The fixed cells were then washed three times with phosphate buffered saline (PBS) (C10010500BT, Gibco) for 1 min each and stained with Oil Red O (G1260, Solarbio) at 37 °C for 1 h. Stained cells were washed twice with 60% isopropanol (1.17029.023, GHTECH) for 1 min each and then washed with PBS, 4 to 5 times for 1 min each. Formation of lipid droplets was observed under a microscope.

Identification and feature analysis of eRNAs in muscle and fat tissues

To identify eRNAs on a genome-wide scale in muscle and fat tissues from both ES and Duroc pigs, first we identified enhancers defined based on H3K27ac signal intensity [ 36 ], in line with the recognized definition that eRNAs are transcribed from enhancer regions. In total, we identified 69,881 enhancers in muscle and fat tissues from ES and Duroc pigs. Among them, 31,336, 27,148, 51,898, and 36,635 enhancers were identified in Duroc muscle, Duroc fat, ES muscle, and ES fat, respectively (see Additional file 6 : Table S4). Samples were clustered based on H3K27ac signal intensity, which showed that enhancer activity was more conserved between breeds than between tissues (Fig.  1 a). Next, based on the coordinates of these enhancers, we used RNA-seq data from the same individuals for which H3K27ac data were used to quantify the expression levels of eRNAs in each respective tissue, referring to the identification methodology used in a previous eRNA study [ 29 ]. Using RNA-seq data, the expression level of eRNAs was quantified with the SeqMonk software, which is referred to as the Nancy method [ 30 ] (see Additional file 7 : Figure S2a). Finally, we identified 17,715 eRNAs, of which 12,430, 14,828, 16,214, and 13,612 were detected in Duroc muscle, Duroc fat, ES muscle, and ES fat, respectively (see Additional file 8 : Table S5). eRNA expression clustering revealed high reproducibility between biological replicates and a higher expression similarity between breeds than between tissues (Fig.  1 c). Although eRNA expression was found across all pig breeds and tissues, a small number of the eRNAs were expressed in a tissue/breed specific pattern (Fig.  1 b and see Additional file 9 : Figure S3). In each tissue, the expression of eRNA is positively correlated with H3K27ac enhancer activity (P < 2.2e−16), which is consistent with prior investigations in this field (Fig.  1 d and see Additional file 10 : Figure S4a–c). We also identified SE in muscle and fat tissues from ES and Duroc pigs (see Additional file 11 : Figure S5a–d). The significantly higher expression of eRNAs within than outside SE (P < 2.2e−16) (Fig.  1 e), underscored the potential of eRNAs to regulate key genes that define cell identities [ 63 , 64 ].

Identification and regulatory characteristics of eRNAs. a A heatmap revealed enhancer intensity dynamics in muscle and adipose tissues between Enshi Black (ES) and Duroc pigs. Enhancer intensity was calculated by normalized H3K27ac ChIP-seq signals. b Overlap of detectable eRNAs between different tissues and breeds. The bar chart in light green displays the number of detectable eRNAs for each tissue, the blue diagram illustrates the degree of overlap between detected eRNAs across different tissues and breeds. c Heatmap displaying the calculated Euclidean distances between samples based on the variance-stabilizing transformation of the eRNA expression matrix. d Pearson correlation between eRNA expression and enhancer activity in Duroc muscle tissue. e Comparison of eRNA expression levels within and outside of super-enhancers. Statistical significance was assessed using the two-sided unpaired Wilcoxon test (** P  < 2.2e−16). f Proportion of detectable unidirectional- and bidirectional-transcribed eRNA per tissue. g Expression levels of unidirectional- and bidirectional-transcribed eRNAs in each tissue. Statistical significance was assessed using the two-sided unpaired Wilcoxon test (** P  < 2.2e−16). h GC content of unidirectional- and bidirectional-transcribed eRNA sequences in each tissue. Statistical significance was assessed using the Student’s t-test (** P  < 2.2e−16). i Deep learning model distinguishes transcriptional directions of eRNAs

eRNA transcription was previously known to be bidirectional, but subsequent studies have shown that not all eRNAs are bidirectionally transcribed [ 65 , 66 ]. Consistently, we found that only about 3/4 of the eRNAs in each tissue were bidirectionally transcribed (Fig.  1 f). Compared to unidirectionally transcribed eRNAs, the bidirectionally transcribed eRNAs have lower expression (Fig.  1 g) and higher GC levels (Fig.  1 h). Furthermore, the adjacent gene sets associated with both unidirectional and bidirectional transcribed eRNAs demonstrate enrichment in distinct biological processes (see Additional file 7 : Figure S2b). In spite of a modest but significant difference in GC content for these two eRNA classes, distinguishing enhancers with distinct transcription directions posed a challenge when relying on higher-level sequence features. Using deep learning methods, including DeepSEA [ 48 ] and LSTM [ 49 ], we observed limited discriminatory power, with AUC values of 0.54 and 0.53, respectively (Fig. 1 i). To the best of our knowledge, this compilation of bidirectional and unidirectional eRNAs represents the first comprehensive eRNA expression profile documented in pig tissues.

The long terminal repeat (LTR) retrotransposon family might promote enhancer transcription

In mammals, transposon sequences account for about half of the entire genome [ 67 ]. They play a significant role in the generation of numerous cis-regulatory elements by incorporating sequence features that are rich in information into the genome, including TF binding motifs [ 68 , 69 ]. In our study, about 81% of the identified enhancers marked by H3K27ac had transposon insertions, which implies that transposons have an important role in the formation of enhancers. However, whether certain types of transposons preferentially contribute to the formation of eRNA loci is unexplored. Following this idea, we classified the enhancers into two categories: transcribed enhancers (TEn) and non-transcribed enhancers (non-TEn). We found that non-TEn (79%) had subtle but significantly lower frequencies of transposon insertion events than TEn (83%) (Fig.  2 a; P  < 2.2e−16). The percentage of transposon bases in TEn was also higher than in non-TEn (Fig.  2 b), and this difference was mainly due to an increased presence of the long terminal repeat (LTR) retrotransposon family in TEn, compared to short (SINE), and long (LINE) interspersed sequences, DNA repeats, or other types of repeats (Fig.  2 c), which indicates a more prominent role of retroviral LTR in the transcription of enhancers. To further study the effect of the LTR retrotransposon family on the transcription of enhancers, we performed enrichment analysis of different transposon families in enhancers. All LTR families were significantly and specifically enriched in TEn compared to non-TEn except for the LTR/ERVK family (Fig.  2 d). The limited enrichment of this specific LTR family in the TEn may be attributed to its infrequent presence within the genome, which subsequently reduces the statistical robustness of the enrichment analysis (Fig.  2 e). In summary, we infer that the insertion of LTR sequences might play an important role in eRNA transcription.

The effect of transposons on enhancer transcription. a Percentages of transposon insertion in transcribed enhancers (TEn) and non-transcribed enhancers (non-TEn). Statistical significance was assessed using Fisher’s exact test. b Proportion of transposon bases within TEn and non-TEn. c Proportion of enhancer occupancy by TE classes. d Enrichment analysis of transposon family in TEn and non-TEn. Statistical significance was determined using a permutation test with genomic background simulation of elements performed 1000 times and adjusted false discovery rate (FDR) (*P < 0.05). e Number of long terminal repeats (LTR) retrotransposon families annotated in the pig genome

Tissue-specific eRNAs exhibit robust regulatory potential in tissue-specific-biological processes

Next, we explored the tissue specificity of eRNAs, as the tissue-specific regulatory landscape is thought to have a substantial impact on trait and disease etiology [ 70 , 71 , 72 ]. Our hypothesis was that eRNAs that are specifically expressed in pig muscle and fat tissues are closely related to pork production traits. To this end, we downloaded additional RNA-seq data from heart, liver and spleen tissues of ES and Duroc pigs (GEO Data repository GSE143288). Then, we analyzed the expression level of the 17,715 eRNAs in five tissues across the two breeds (see Additional file 12 : Table S6). The heatmap clustering analysis revealed a striking consistency in eRNA expression patterns across these two breeds within the same tissue, quantified in reads per million (RPM). While most biological replicates displayed cohesive clustering, it is noteworthy that, in one case, a replicate of skeletal muscle tissue from the ES breed clustered with ES heart tissue. This deviation could be ascribed to the underlying physiological and functional similarities between the skeletal and cardiac muscles [ 73 ] (see Additional file 9 : Figure S3). Interestingly, this similarity in eRNA expression within tissues transcended breed variations. Conversely, when examining a particular breed across different tissues, the eRNA expression patterns displayed more divergence. This analysis underscores the presence of tissue-specific eRNAs, which implies that they play vital roles in orchestrating tissue-specific regulatory processes.

Next, we used the tissue specificity index (TSI) algorithm to identify tissue-specific eRNAs in each tissue of each breed (see “ Methods ” for details). In total, we identified 643, 1188, 870, and 670 eRNAs specifically expressed in Duroc muscle, Duroc fat, ES muscle, and ES fat tissues, respectively (Fig.  3 a and see Additional file 13 : Table S7). The genes that were localized nearest to these tissue-specific eRNAs were significantly enriched in biological pathways that are involved in the differentiation and development of the corresponding tissues (Fig.  3 b). Although the neighboring genes of Duroc fat-specific eRNAs did not exhibit direct enrichment in pathways specifically associated with fat deposition, they displayed enrichment in pathways relevant to adipogenesis, such as the Wnt signaling pathway [ 74 , 75 ]. However, the neighboring genes of non-tissue-specific eRNAs exhibited a remarkable pattern of enrichment in essential biological pathways that are shared across diverse tissues (see Additional file 14 : Figure S6). These GO analyses suggested that tissue-specific eRNAs play an important role in regulating biological processes that are relevant to tissue identity. With the exception of Duroc adipose tissue ( P  = 0.056), we found that tissue-specific eRNAs were significantly enriched within SE of the corresponding tissues (Fig.  3 c). Furthermore, we identified 421, 506, 594 and 412 genes that were specifically expressed in Duroc muscle, Duroc fat, ES muscle and ES fat tissues, respectively (Fig.  3 a and see Additional file 13 : Table S7). In alignment with the adjacent genes of tissue-specific eRNAs, tissue-specific genes were also significantly enriched in biological pathways that are relevant to tissue identity, except for Duroc fat (see Additional file 15 : Figure S7). The functional relevance of the tissue-specific eRNAs and genes led us to speculate that tissue-specific genes are primarily regulated by tissue-specific eRNAs. To explore the potential regulatory relationship between tissue-specific eRNAs and tissue-specific genes, we conducted an enrichment analysis comparing tissue-specific genes to tissue-specific eRNAs within a ± 1 Mb region, which is a commonly employed criterion for assigning putative target genes to eRNAs [ 27 , 31 ]. Our findings revealed a significant enrichment of tissue-specific gene locations in tissue-specific eRNAs within ± 1 Mb, in the same tissue (Fig.  3 d). This enrichment implies that tissue-specific eRNAs may have robust regulatory activities on the expression of tissue-specific genes. To investigate which TF might be involved in the regulation of tissue-specific eRNAs, we performed TF binding motif enrichment analysis of tissue-specific eRNAs. As expected, the top enriched motifs in tissue-specific eRNAs are DNA binding motifs of master TF that are closely related to tissue development. For example, MYOG, MYOD, MYF5, MEF2A, and MEDF2D were significantly enriched in muscle-specific eRNAs, while ATF3 [ 76 , 77 ], BACH1 [ 78 ], and STAT1 [ 79 ] were significantly enriched in fat-specific eRNAs (Fig.  3 e; Q-value < 0.01). Therefore, tissue-specific eRNAs are likely to function downstream of these master TF to establish and maintain tissue identity. Accordingly, DNA variants that dysregulate these eRNAs might significantly contribute to the genetic components underlying pork economic traits relevant to muscle and fat.

Analysis of the biological functions and the potential regulatory mechanisms of tissue-specific eRNAs. a The number of tissue-specific eRNAs and genes in each tissue between Enshi Black (ES) and Duroc pigs. b Gene ontology (GO) analysis reveals biological process pathways relevant to eRNAs expressed in a tissue- and breed-specific manner. GO enrichment analysis was performed based on the neighboring genes of eRNAs. c Tissue-specific genes were significantly enriched within ± 1 Mb distance of tissue-specific eRNAs. Statistical significance was assessed using the hypergeometric test (**P < 0.01). d Enrichment analysis of tissue-specific eRNAs in super-enhancers (SE). Statistical significance was assessed using the hypergeometric test (**P < 0.01). e Motif enrichment analysis of tissue-specific eRNAs. Red dots represent transcription factors (TF) with a DNA binding motif that were significantly enriched in adipose-specific eRNAs; green dots represent those that were significantly enriched in muscle-specific eRNAs; dark blue dots represent those that were significantly enriched in both fat and muscle-specific eRNAs. The size of the dots represents the expression level of corresponding TF

Genetic contribution of tissue-specific eRNAs to tissue-relevant traits

Numerous post-GWAS studies have dissected the heritability of complex traits by fine-mapping DNA variants in the non-coding regions of genomes [ 80 , 81 ]. Thus, we postulate that tissue-specific eRNAs might provide a unique resource to functionally interpret GWAS signals that lie outside of coding regions, and thereby pinpointing novel causal regulatory elements and genes. To explore how muscle- and fat-specific eRNAs might shed light on the genetic basis of muscle- and fat-relevant pork production traits, we first collected all GWAS hits and QTL associated with different porcine traits from the pig QTLdb [ 82 ]. These traits are categorized into five major groups (Meat and Carcass, Health, Production, Reproduction, Exterior) in the pigQTLdb database. We observed a significant and specific enrichment of muscle- and adipose-specific eRNAs within 20-kb linkage regions in the GWAS signals related to meat and carcass traits, as well as to production traits (Fig.  4 a). These two groups of traits are largely affected by muscle and fat development. In the enrichment analysis using QTL regions, muscle- and fat-specific expression eRNAs were not only enriched in muscle and fat development-related traits (Fig.  4 b), but also in reproductive and health traits. This lack of enrichment specificity might be due to the fact that the pig QTL regions were generally much longer than the GWAS regions.

Contribution of tissue-specific eRNAs for the genetic basis of pig economic traits. a Enrichment of genome-wide association study (GWAS) signals of five major pig economic traits in tissue-specific eRNAs. A permutation test was performed with genomic background simulation of eRNA elements conducted 1000 times and a false discovery rate (FDR) adjusted for statistical significance (*P < 0.05; **P < 0.01). b Enrichment of tissue-specific eRNAs in trait-related quantitative trait loci (QTL) regions. Statistical significance was determined using a permutation test with an adjusted FDR (*P < 0.05; **P < 0.01). c Number of pig tissue-specific eRNAs and their homologs in the human genome. d A heatmap displaying stratified linkage disequilibrium score regression (S-LDSC) heritability enrichment of 64 human GWAS traits in human homologs of pig tissue-specific eRNAs. Data are hierarchically clustered by GWAS and tissues. * represents significant enrichment (P < 0.05). e A dot plot displaying the statistical significance of S-LDSC enrichment for waist-hip ratio. A red dashed line is marked at −log10(0.05)

To further explore the biological importance of muscle- and fat-specific eRNAs, we took advantage of the human GWAS, which usually provide sharper GWAS signals, and evaluated how the human orthologs of these tissue-specific eRNAs contribute to various human traits. The tissue-specific eRNAs exhibit a high level of sequence homology between pigs and humans, and the majority of them could be successfully lifted over to the human genome (Fig.  4 c). In the comparison between our results and the summary statistics data from 64 GWAS for human traits collected by Hook and McCallion [ 37 ], the stratified linkage disequilibrium score regression (S-LDSC) analysis revealed that the human orthologs of pig muscle- and fat-specific eRNAs were significantly enriched in waist-hip ratio GWAS associated loci (Fig. 4 d, e and see Additional file 16 : Table S8). This finding is consistent with the notion that the regulation of waist-hip ratio involves coordinated mechanisms that govern muscle growth and fat deposition. Thus, we provided evidence that pig muscle- and fat-specific eRNAs might constitute highly conserved transcriptional regulatory networks that underlie crucial muscle- and fat-relevant phenotypes across pigs and humans.

Construction of eRNA-mediated gene regulatory networks underlying muscle- and fat-relevant traits

To pinpoint tissue-specific eRNAs that potentially influence meat-relevant traits, we present a novel approach to construct gene regulatory networks. This innovative method enables the identification of eRNAs, which harbor DNA variants that are likely to contribute to phenotypic variation. Compared to previous co-expression-based gene regulatory networks (GRN) that link genes with undirected edges [ 83 ], our networks (referred to as eGRN) are eRNA-centric, have directed edges and leverage information from population genetics studies. Our methodology involves several key steps. First, we used co-expression analysis and motif scanning to associate upstream TF with tissue-specific eRNAs. Next, we used a combination of co-expression patterns, a 1-Mb distance cutoff, and topologically associated domain (TAD) boundaries to link these eRNAs to potential downstream target genes. Subsequently, we leveraged pig GWAS data to keep only the gene regulatory networks mediated by eRNAs that harbor DNA variants exhibiting potential associations with meat-relevant traits (Fig.  5 a, and see “ Methods ”).

Construction of potential trait-affecting eRNA-mediated gene regulatory networks (eGRN) by integrating multi-omics data. a Schematic diagram of the eGRN construction method. Motif scanning and expression correlation between transcription factors (TF) and eRNA expression levels were used to predict upstream TF of eRNAs. The distance between eRNAs and genes and the expression correlation between their expression levels were used to predict downstream target genes of eRNAs, topologically associated domain (TAD) structure was further used to filter out eRNAs and genes that were not in a same TAD. Genome-wide association study (GWAS) data were used to further filter for trait-associated hub eRNAs that harbor DNA variants that are associated with pig economic traits. b Heat maps showing the globally correlated changes of the expression of eRNAs and target genes. c eGRN for muscle- and fat-related traits. The color of a TF represents the number of its downstream regulatory eRNAs, where a darker color indicates a larger number of regulatory eRNAs. d Differential expression of eRNA target genes between ES and Duroc pigs (ES vs Duroc) in fat-related eGRN

Based on our approach, 61% of the tissue-specific expression eRNAs were assigned to at least one target gene and most of the eRNAs were assigned to no more than five (1680/2751) (see Additional file 17 : Figure S8A). eRNAs and their target genes showed notable and coordinated expression changes, which suggest a robust and synergistic relationship between them (Fig.  5 b). Gene ontology enrichment analyses revealed that although the target genes of Duroc fat- and ES muscle-specific eRNAs were not enriched in tissue-related pathways, those of Duroc muscle- and ES fat-specific eRNAs were both significantly enriched in tissue-related biological pathways (see Additional file 17 : Figure S8b). This pattern is reminiscent of the selection for muscle mass and fat deposition in Duroc and ES during pig breeding practice, respectively, indicating distinct regulatory roles of the tissue-specific eRNAs that underlie selection during the domestication and breed development of pigs. The eGRN of muscle and fat tissues encompass 1023 and 711 regulatory relationships, respectively, involving upstream TF binding to eRNA-transcribing enhancers for the regulation of downstream target genes (Fig.  5 c and see Additional file 18 : Table S9). Excitingly, we found that the upstream TF of these eRNAs include many known master regulators for tissue development, such as MYOG, MYOD, MYF5, MEF2A, and MEDF2D in the muscle eGRN, and CREB5, IRF4, EBF1, EBF2, KLF6, KLF4, GATA2, GATA3, KLF14, and STAT6 in the fat eGRN [ 84 , 85 ]. In addition, the target genes of the eGRN also include genes that have been reported to regulate tissue development, such as MYOD1 , MYH1 , MYH2 in the muscle eGRN, as well as SCD and SLC27A6 in the fat eGRN. Interestingly, the expression level of MYOD1, MYH1, MYH2 was higher in Duroc muscle than in ES muscle, while the expression level of SCD and SLC27A6 was higher in ES fat than in Duroc fat (Fig.  5 d and see Additional file 17 : Figure S8c). This difference in gene expression is consistent with differences in lean mass ratio and fat deposition between Duroc and ES pigs. Taken together, these results provide compelling support for the reliability of our eGRN in capturing essential transcriptional regulatory units that underlie relevant traits.

Unraveling novel regulators of adipocyte differentiation by refining eGRN with STARR-Seq

It is important to note that GWAS associations between genetic loci and traits are often statistical in nature, rather than indicative of causal relationships. Thus, fine-mapping trait-associated genetic loci is pivotal for unraveling the mechanisms of trait inheritance. In our study, we constructed an eGRN through eRNA-transcribing regions that harbor DNA variants with potential pig GWAS signals. To enhance the precision of the eGRN, we subsequently performed capture STARR-Seq experiments, enabling high-throughput screening of enhancer activity (Fig.  6 a). Through these experiments, we refined the eGRN by selectively retaining eRNA regions that demonstrated allele-specific enhancer activity. Taking the fat trait-associated eGRN as an example, we exploited a series of criteria to select SNPs within eRNA-transcribing regions for STARR-seq validation. These criteria encompassed SNPs that were detected within the pooled population of eastern and western pigs, characterized by a minimum difference in minor allele frequency (MAF) of 0.3 between the two populations, as well as a MAF exceeding 0.05 in the mixed pool. After applying these criteria (see Additional file 4 : Figure S1b), we selected 107 SNPs for STARR-seq assays (see Additional file 19 : Table S10), and identified 16 regulatory SNPs that influenced enhancer activity in 3T3 cells (Fig.  6 b and see Additional file 20 : Table S11). We designated the eRNAs with these regulatory SNPs as hub eRNAs and used them to identify refined eGRN that might play crucial roles in regulating fat-related economic traits (see Additional file 21 : Table S12). Within the refined eGRN, we observed 24 potential target genes that could be regulated by eRNAs (Fig.  6 c).

Refining eRNA-mediated gene regulatory networks (eGRN) by self-transcribing active regulatory region sequencing (STARR-Seq) and subsequent identification of functional genes in adipocyte differentiation. a The workflow for STARR-seq strategy to detect SNPs influencing DNA activity. b Regulatory SNPs identified by STARR-seq that affect enhancer activity. c Refined eGRN underlying fat-related traits identified through regulatory SNP screening. d siRNA knockdown effects of five potential target genes on adipocyte differentiation assessed by Oil Red O Staining in 3T3-L1 pre-adipocytes. NC represents the blank control. e siRNA knockdown effects of potential target genes on marker gene expression during adipocyte differentiation in 3T3-L1 pre-adipocytes. (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001)

To confirm the regulatory role of the refined eGRN on fat-related traits, we performed RNA interference experiments on five randomly selected target genes ( SLC27A6 , RB1 , APLF , RGMA , UBTD2 ) in the widely used pre-adipocyte 3T3-L1 cell line derived from mouse embryos, which serves as a popular model for investigating adipocyte differentiation. The mouse Pparg , Cebpa , and Lpl genes, which are widely recognized as master regulators of adipogenesis [ 86 , 87 , 88 , 89 ], were used as marker genes to assess the impact of the target genes on adipogenesis. Among the five potential target genes, we found that APLF , RGMA , and UBTD2 played a negative role in adipogenesis while SLC27A6 and RB1 played a positive role (Fig.  6 e). Oil staining experiments revealed consistent results (Fig.  6 d). It is noteworthy that the APLF , RGMA , and UBTD2 genes exhibited higher expression levels in western Duroc pigs that are characterized by high lean meat mass. By comparison, the SCL27A6 and RB1 genes showed higher expression levels in eastern ES pigs that are known for their superior fat deposition capacity (Fig.  5 d). Therefore, the results of the in vitro experimental validation are highly consistent with the correlation between gene expression and meat-relevant phenotypes, indicating that these genes hold promising potential as important candidate genes for future improvement of pork quality. To summarize, we refined fat eGRN by STARR-seq assays, and further validated the functionality of five genes in adipocyte differentiation using RNA interference experiments. Our findings not only provide potential molecular markers for accelerating pig breeding programs, but also offer a novel eRNA-centric approach for fine-mapping GWAS signals and constructing crucial gene regulatory networks that underlie complex traits.

eRNAs are emerging as a type of critical players in transcriptional regulation, which means that they are particularly relevant for deciphering the genetic basis of complex traits and diseases. In spite of substantial efforts to map eRNAs in various biological contexts, there has been limited research on building eRNA-centric transcriptional regulatory networks and using these networks to pinpoint important genes that potentially affect phenotypes. In this study, by combining whole-transcriptome RNA-Seq-based eRNA profiling, GWAS signals and high-throughput STARR-Seq experiments, we present a novel approach to construct trait-affecting hub eGRN and to identify potential trait-affecting genes. We demonstrate the utility of this approach in dissecting the genetic basis of pork production traits and propose potential fat deposition-affecting genes with evidence on their role in adipocyte differentiation in a cell line model. In light of the exponential growth in whole-transcriptome RNA-Seq data across diverse biological contexts, the ever-expanding repertoire of GWAS loci identified by large-scale population genetic studies (such as UK Biobank), and the rising utilization of STARR-Seq (or other MPRA-based techniques) for studying the cis-regulatory impacts of non-coding DNA variants, we believe that our approach possesses great potential for extensively unraveling the genetic basis underlying a wide spectrum of complex traits and diseases across diverse species.

Biological systems involve a diversity of regulatory processes. To understand the molecular bases that underlie these processes, GRN have been extensively built by different experimental and computational methods. Here, we present a novel approach to construct GRN based on tissue-specific eRNAs that have substantial implications in the formation of muscle and fat-related traits. Compared to conventional gene regulatory networks, our eGRN have directionality, encompassing the intricate regulatory cascade that unfolds from upstream TF to eRNAs, and subsequently extending to downstream target genes. Using muscle and fat as examples, we showed that in our eGRN, many upstream TF of eRNAs are known major regulators that affect tissue development, and the functional annotation of downstream target genes exhibited close relevance to tissue-related biological processes. Notably, the precision and robustness of the eGRN were further increased after we combined the GWAS data and STARR-seq data. As an example, through the identification of eRNAs that harbor cis-regulatory SNPs based on pig GWAS signals, we uncovered hub eRNAs and their potential target genes that influence adipocyte differentiation in 3T3-L1 cells. Although our approach for network construction has proven effective and indicates general applicability, it is important to acknowledge its inherent limitations. For instance, when identifying target genes of eRNAs, we typically refer to previous research [ 27 , 31 ], which suggests that potential target genes of eRNAs are generally located within a ± 1 Mb region of the eRNA. Nevertheless, distal enhancers persist in regulating gene expression even when located more than ± 1 Mb away from transcription start sites (TSS). Refining the identification of enhancer-gene pairs stands to gain significant advantages from technologies capable of assaying chromatin interactions, such as Hi-C and HiChIP. in addition, employing motif sequence scanning to infer TF binding and the incorporation of TAD constraints to predict downstream target genes, while grounded in fundamental biological principles, may also inadvertently introduce noise into the regulatory relationships within the network. Given the ongoing efforts in producing various types of epigenetic data in a wide range of biological contexts, we anticipate that the integration of transcription factor ChIP-seq data and high-order chromatin structure Hi-C data will further optimize our network construction approach, leading to improved accuracy in identifying both upstream TF and downstream target genes of eRNAs.

Recent years have seen a surge of eRNA-related studies, which have identified eRNAs using high-depth RNA-seq data [ 28 , 29 , 30 , 31 , 90 ]. The low and unstable expression of eRNAs makes high-depth RNA-seq data critical for accurate eRNA identification. In our study, the library size of the RNA-seq data ranged from 38 to 50 million reads per sample, providing a higher depth of RNA-seq data for eRNA identification compared to the study of Carullo et al. [ 30 ]. Furthermore, the RNA-seq samples displayed strong reproducibility (Fig.  1 c), enabling us to accurately profile an eRNA expression atlas in pigs. The previous understanding of eRNAs was that they were transcribed in a bi-directional manner [ 91 ], However, subsequent studies have revealed that some eRNAs are transcribed in a unidirectional fashion [ 65 , 66 ], which is supported by the results of our study, where we identified different transcriptional patterns of eRNA in tissues and discovered that they have different molecular characteristics and are involved in different biological processes. Although deep learning classification models are used for sequence feature learning, the classification of unidirectional and bidirectional transcribed enhancers remains a challenging endeavor. This ongoing difficulty suggests that the transcription direction of eRNAs may not be only determined by the linear sequence itself, but rather that it may be influenced by complex higher-order chromatin structural dynamics that extend beyond the scope of eRNA sequence analysis.

Transposable elements play a crucial role in shaping the evolution of cis-regulatory elements. An intriguing finding in our study is the significant and specific enrichment of LTR families in TEn. However, it is worth noting that the LTR/ERVK family did not exhibit significant enrichment, which is likely due to insufficient statistical power as a result of its small number of copies in the pig genome. Previous research has established the role of LTR transposons in driving the transcription of lncRNAs in the human and mouse genomes [ 92 ]. As a result, we propose that LTR transposons might also play a key role in the transcription of eRNA.

Over the years, there have been many studies on the genetic basis of pig production traits. For example, candidate SNP or genes that affect pig production traits have been identified through GWAS [ 93 , 94 , 95 ], by differential gene expression analysis or co-expression network analysis using RNA-seq data from different pig breeds [ 96 , 97 , 98 ], and by integrating epigenetic modifications [ 99 , 100 ]. While these studies have undoubtedly offered valuable insights into the genetic mechanisms that underlie pork production traits, it is crucial to recognize that biological processes operate through intricate, multi-layered networks. Relying solely on the analysis of individual or limited omics data is inadequate to comprehensively unravel the complex genetic regulatory mechanisms that govern these traits. Aligned with this perspective, our study exemplifies an integrative approach that demonstrates the power of exploiting multi-omics data to dissect the genetic basis of pig economic traits.

The functional annotation of genomes has provided a solid foundation for the interpretation of regulatory mechanisms of genetic variations, particularly those in the non-coding regions [ 101 , 102 ]. Previous studies have shown that tissue-specific cis-regulatory elements enriched for GWAS signals are associated with complex phenotypes [ 72 ], highlighting the importance of this type of functional annotation in understanding the genetic basis of phenotypic variations. Here, by comparing eRNA annotation with pig GWAS and QTL data, our results suggest the pivotal role of tissue-specific eRNAs in shaping economic traits in pigs. Beyond pig complex traits, our study also demonstrated that pig muscle- and fat-specific eRNAs can effectively serve as surrogates to decipher the hereditary nature of human muscle- and fat-related phenotypes. This finding further strengthens the utilization of pigs as models for medical research in humans. This is reminiscent of using orthologous-based mouse-derived human open chromatin profiles for understanding the genetic basis of human diseases or phenotypes [ 37 ]. Notably, compared to other model species, pigs are generally regarded as anatomically and developmentally more similar to humans, enhancing their relevance as model organisms.

While our systematic study on pig tissue eRNAs provides a valuable approach for understanding complex traits in pigs, it still has certain limitations. For instance, in the identification of eRNAs, we followed the methodology outlined by Carullo et al. [ 30 ], excluding known coding gene and non-coding RNA regions based on the reference genome annotation to prevent interference with enhancer transcription signals. In spite of our efforts to leverage annotated genomic files from different databases (RefSeq, UCSC, and Ensembl) for the pig reference genome, the possibility of incomplete annotation remains. Unannotated genes or transcription elements in the pig reference genome are an inevitable objective reality. In addition, when our aim was to screen for eRNAs associated with GWAS hits linked to specific traits, we referred to methodologies applied in previous studies [ 103 , 104 ], using a 20-kb window to consider the effect of LD of SNPs instead of directly computing LD values between SNPs. However, we recognize that under conditions where direct LD calculations are feasible for the pig population, selecting trait-associated SNPs within eRNAs could yield more precise results. Furthermore, although our study provides the first pig tissue-level eRNA expression atlas and offers a framework for constructing eGRN to dissect meat-related traits in pigs, there is room for optimization. Our focus was on 2-week-old pigs from eastern and western breeds, which overlooked the temporal specificity of eRNA expression at various developmental stages. Future integration of complementary transcriptomic and H3K27ac data across additional developmental time points will enhance the utility of eGRN in unraveling the genetic mechanisms for meat-related traits in pigs.

In summary, our work not only identified critical eRNAs and their target genes that might contribute to the genetic basis of pork production traits, but also proposed a novel strategy to construct eRNA-centered GRN that pinpoint potential trait-affecting genes. This strategy is broadly applicable for uncovering how genetic components in cis-regulatory elements, such as enhancers, might shape phenotypic diversity in various organisms.

Availability of data and materials

The raw sequence data of STARR-seq used in this paper have been deposited (PRJCA017321) in the Genome Sequence Archive in the BIG Data Center ( http://bigd.big.ac.cn/ ) with the accession code CRA011292.

Edwards SL, Beesley J, French JD, Dunning AM. Beyond GWASs: illuminating the dark road from association to function. Am J Hum Genet. 2013;93:779–97.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Mountjoy E, Schmidt EM, Carmona M, Schwartzentruber J, Peat G, Miranda A, et al. An open approach to systematically prioritize causal variants and genes at all published human GWASs trait-associated loci. Nat Genet. 2021;53:1527–33.

Farh KK, Marson A, Zhu J, Kleinewietfeld M, Housley WJ, Beik S, et al. Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature. 2015;518:337–43.

Article   CAS   PubMed   Google Scholar  

Mouse ENCODE Consortium, Stamatoyannopoulos JA, Snyder M, Hardison R, Ren B, Gingeras T, et al. An encyclopedia of mouse DNA elements (mouse ENCODE). Genome Biol. 2012;13:418.

Article   Google Scholar  

ENCODE Project Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature. 2012;489:57–74.

Andersson L, Archibald AL, Bottema CD, Brauning R, Burgess SC, Burt DW, et al. Coordinated international action to accelerate genome-to-phenome with FAANG, the Functional Annotation of Animal Genomes project. Genome Biol. 2015;16:57.

Article   PubMed   PubMed Central   Google Scholar  

Miguel-Escalada I, Bonas-Guarch S, Cebola I, Ponsa-Cobas J, Mendieta-Esteban J, Atla G, et al. Human pancreatic islet three-dimensional chromatin architecture provides insights into the genetics of type 2 diabetes. Nat Genet. 2019;51:1137–48.

Cooper YA, Teyssier N, Drager NM, Guo Q, Davis JE, Sattler SM, et al. Functional regulatory variants implicate distinct transcriptional networks in dementia. Science. 2022;377:eabi8654.

Chella Krishnan K, Kurt Z, Barrere-Cain R, Sabir S, Das A, Floyd R, et al. Integration of multi-omics data from mouse diversity panel highlights mitochondrial dysfunction in non-alcoholic fatty liver disease. Cell Syst. 2018;6:103-115.e7.

Liu Y, Fu Y, Yang Y, Yi G, Lian J, Xie B, et al. Integration of multi-omics data reveals cis-regulatory variants that are associated with phenotypic differentiation of eastern from western pigs. Genet Sel Evol. 2022;54:62.

Li J, Xiang Y, Zhang L, Qi X, Zheng Z, Zhou P, et al. Enhancer-promoter interaction maps provide insights into skeletal muscle-related traits in pig genome. BMC Biol. 2022;20:136.

Gondor A, Ohlsson R. Enhancer functions in three dimensions: beyond the flat world perspective. F1000 Res. 2018;7:681.

Han Z, Li W. Enhancer RNA: what we know and what we can achieve. Cell Prolif. 2022;55: e13202.

Mikhaylichenko O, Bondarenko V, Harnett D, Schor IE, Males M, Viales RR, et al. The degree of enhancer or promoter activity is reflected by the levels and directionality of eRNA transcription. Genes Dev. 2018;32:42–57.

Sanyal A, Lajoie BR, Jain G, Dekker J. The long-range interaction landscape of gene promoters. Nature. 2012;489:109–13.

Kaikkonen MU, Spann NJ, Heinz S, Romanoski CE, Allison KA, Stender JD, et al. Remodeling of the enhancer landscape during macrophage activation is coupled to enhancer transcription. Mol Cell. 2013;51:310–25.

Arnold PR, Wells AD, Li XC. Diversity and emerging roles of enhancer RNA in regulation of gene expression and cell fate. Front Cell Dev Biol. 2019;7:377.

Article   PubMed   Google Scholar  

Wang Y, Zhang C, Wang Y, Liu X, Zhang Z. Enhancer RNA (eRNA) in human diseases. Int J Mol Sci. 2022;23:11582.

Core LJ, Waterfall JJ, Lis JT. Nascent RNA sequencing reveals widespread pausing and divergent initiation at human promoters. Science. 2008;322:1845–8.

Kwak H, Fuda NJ, Core LJ, Lis JT. Precise maps of RNA polymerase reveal how promoters direct initiation and pausing. Science. 2013;339:950–3.

Churchman LS, Weissman JS. Native elongating transcript sequencing (NET-seq). Curr Protoc Mol Biol. 2012;Chapter 4:Unit4.14.1-17.

Google Scholar  

Mejia-Guerra MK, Li W, Galeano NF, Vidal M, Gray J, Doseff AI, et al. Core promoter plasticity between maize tissues and genotypes contrasts with predominance of sharp transcription initiation sites. Plant Cell. 2015;27:3309–20.

De Santa F, Barozzi I, Mietton F, Ghisletti S, Polletti S, Tusi BK, et al. A large fraction of extragenic RNA pol II transcription sites overlap enhancers. PLoS Biol. 2010;8: e1000384.

Djebali S, Davis CA, Merkel A, Dobin A, Lassmann T, Mortazavi A, et al. Landscape of transcription in human cells. Nature. 2012;489:101–8.

Murakawa Y, Yoshihara M, Kawaji H, Nishikawa M, Zayed H, Suzuki H, et al. Enhanced identification of transcriptional enhancers provides mechanistic insights into diseases. Trends Genet. 2016;32:76–88.

Wu H, Nord AS, Akiyama JA, Shoukry M, Afzal V, Rubin EM, et al. Tissue-specific RNA expression marks distant-acting developmental enhancers. PLoS Genet. 2014;10: e1004610.

Zhang Z, Hong W, Ruan H, Jing Y, Li S, Liu Y, et al. HeRA: an atlas of enhancer RNAs across human tissues. Nucleic Acids Res. 2021;49:D932–8.

Chen H, Li C, Peng X, Zhou Z, Weinstein JN, Cancer Genome Atlas Research Network, et al. A pan-cancer analysis of enhancer expression in nearly 9000 patient samples. Cell. 2018;173:386–99.

Zhang Z, Lee JH, Ruan H, Ye Y, Krakowiak J, Hu Q, et al. Transcriptional landscape and clinical utility of enhancer RNAs for eRNA-targeted therapy in cancer. Nat Commun. 2019;10:4562.

Carullo NV, Phillips RA, Simon RC, Roman Soto SA, Hinds JE, Salisbury AJ, et al. Enhancer RNAs predict enhancer-gene regulatory links and are critical for enhancer function in neuronal systems. Nucleic Acids Res. 2020;48:9550–70.

Jin W, Jiang G, Yang Y, Yang J, Yang W, Wang D, et al. Animal-eRNAdb: a comprehensive animal enhancer RNA database. Nucleic Acids Res. 2022;50:D46-53.

Li M, Chen L, Tian S, Lin Y, Tang Q, Zhou X, et al. Comprehensive variation discovery and recovery of missing sequence in the pig genome using multiple de novo assemblies. Genome Res. 2017;27:865–74.

Ai H, Fang X, Yang B, Huang Z, Chen H, Mao L, et al. Adaptation and possible ancient interspecies introgression in pigs identified by whole-genome sequencing. Nat Genet. 2015;47:217–25.

Arnold CD, Gerlach D, Stelzer C, Boryn LM, Rath M, Stark A. Genome-wide quantitative enhancer activity maps identified by STARR-seq. Science. 2013;339:1074–7.

Liu S, Liu Y, Zhang Q, Wu J, Liang J, Yu S, et al. Systematic identification of regulatory variants associated with cancer risk. Genome Biol. 2017;18:194.

Zhao Y, Hou Y, Xu Y, Luan Y, Zhou H, Qi X, et al. A compendium and comparative epigenomics analysis of cis-regulatory elements in the pig genome. Nat Commun. 2021;12:2217.

Hook PW, McCallion AS. Leveraging mouse chromatin data for heritability enrichment informs common disease architecture and reveals cortical layer contributions to schizophrenia. Genome Res. 2020;30:528–39.

Quinlan AR, Hall IM. Bedtools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26:841–2.

Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9:357–9.

Tarasov A, Vilella AJ, Cuppen E, Nijman IJ, Prins P. Sambamba: fast processing of NGS alignment formats. Bioinformatics. 2015;31:2032–4.

Ramirez F, Dundar F, Diehl S, Gruning BA, Manke T. deepTools: a flexible platform for exploring deep-sequencing data. Nucleic Acids Res. 2014;42:W187–91.

Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 2008;9:R137.

Loven J, Hoke HA, Lin CY, Lau A, Orlando DA, Vakoc CR, et al. Selective inhibition of tumor oncogenes by disruption of super-enhancers. Cell. 2013;153:320–34.

Hah N, Murakami S, Nagari A, Danko CG, Kraus WL. Enhancer transcripts mark active estrogen receptor binding sites. Genome Res. 2013;23:1210–23.

Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015;12:357–60.

Conway JR, Lex A, Gehlenborg N. UpSetR: an R package for the visualization of intersecting sets and their properties. Bioinformatics. 2017;33:2938–40.

Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16:284–7.

Zhou J, Troyanskaya OG. Predicting effects of noncoding variants with deep learning-based sequence model. Nat Methods. 2015;12:931–4.

Min X, Zeng W, Chen N, Chen T, Jiang R. Chromatin accessibility prediction via convolutional long short-term memory networks with k-mer embedding. Bioinformatics. 2017;33:i92-101.

Sachs MC. plotROC: a tool for plotting ROC curves. J Stat Softw. 2017;79:2.

Tarailo-Graovac M, Chen N. Using RepeatMasker to identify repetitive elements in genomic sequences. Curr Protoc Bioinform. 2009;Chapter 4:4.10.1-4.10.14.

Chen N. Using RepeatMasker to identify repetitive elements in genomic sequences. Curr Protoc Bioinform. 2004;Chapter 4:Unit 4.10.

Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30:923–30.

Yanai I, Benjamin H, Shmoish M, Chalifa-Caspi V, Shklar M, Ophir R, et al. Genome-wide midrange transcription profiles reveal expression level relationships in human tissue specification. Bioinformatics. 2005;21:650–9.

Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol Cell. 2010;38:576–89.

Kuhn RM, Haussler D, Kent WJ. The UCSC genome browser and associated tools. Brief Bioinform. 2013;14:144–61.

Finucane HK, Bulik-Sullivan B, Gusev A, Trynka G, Reshef Y, Loh PR, et al. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat Genet. 2015;47:1228–35.

Dixon JR, Selvaraj S, Yue F, Kim A, Li Y, Shen Y, et al. Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature. 2012;485:376–80.

Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13:2498–504.

Zhou ZY, Li A, Otecko NO, Liu YH, Irwin DM, Wang L, et al. PigVar: a database of pig variations and positive selection signatures. Database (Oxford). 2017;2017:bax048.

Duan YY, Chen XF, Zhu RJ, Jia YY, Huang XT, Zhang M, et al. High-throughput functional dissection of noncoding SNPs with biased allelic enhancer activity for insulin resistance-relevant phenotypes. Am J Hum Genet. 2023;110:1266–88.

Liu Y, Yu S, Dhiman VK, Brunetti T, Eckart H, White KP. Functional assessment of human enhancer activities using whole-genome STARR-sequencing. Genome Biol. 2017;18:219.

Hnisz D, Abraham BJ, Lee TI, Lau A, Saint-Andre V, Sigova AA, et al. Super-enhancers in the control of cell identity and disease. Cell. 2013;155:934–47.

Chen C, Zhou D, Gu Y, Wang C, Zhang M, Lin X, et al. SEA version 30: a comprehensive extension and update of the super-enhancer archive. Nucleic Acids Res. 2020;48:D198-203.

CAS   PubMed   Google Scholar  

Koch F, Fenouil R, Gut M, Cauchy P, Albert TK, Zacarias-Cabeza J, et al. Transcription initiation platforms and GTF recruitment at tissue-specific enhancers and promoters. Nat Struct Mol Biol. 2011;18:956–63.

Li W, Notani D, Ma Q, Tanasa B, Nunez E, Chen AY, et al. Functional roles of enhancer RNAs for oestrogen-dependent transcriptional activation. Nature. 2013;498:516–20.

van de Lagemaat LN, Landry JR, Mager DL, Medstrand P. Transposable elements in mammals promote regulatory variation and diversification of genes with specialized functions. Trends Genet. 2003;19:530–6.

Fueyo R, Judd J, Feschotte C, Wysocka J. Roles of transposable elements in the regulation of mammalian transcription. Nat Rev Mol Cell Biol. 2022;23:481–97.

Bourque G, Leong B, Vega VB, Chen X, Lee YL, Srinivasan KG, et al. Evolution of the mammalian transcription factor binding repertoire via transposable elements. Genome Res. 2008;18:1752–62.

Fang L, Cai W, Liu S, Canela-Xandri O, Gao Y, Jiang J, et al. Comprehensive analyses of 723 transcriptomes enhance genetic and biological interpretations for complex traits in cattle. Genome Res. 2020;30:790–801.

Marbach D, Lamparter D, Quon G, Kellis M, Kutalik Z, Bergmann S. Tissue-specific regulatory circuits reveal variable modular perturbations across complex diseases. Nat Methods. 2016;13:366–70.

Pan Z, Yao Y, Yin H, Cai Z, Wang Y, Bai L, et al. Pig genome functional annotation enhances the biological interpretation of complex traits and human disease. Nat Commun. 2021;12:5848.

Lindskog C, Linne J, Fagerberg L, Hallstrom BM, Sundberg CJ, Lindholm M, et al. The human cardiac and skeletal muscle proteomes defined by transcriptomics and antibody-based profiling. BMC Genomics. 2015;16:475.

Song K, Wang S, Mani M, Mani A. Wnt signaling, de novo lipogenesis, adipogenesis and ectopic fat. Oncotarget. 2014;5:11000–3.

Christodoulides C, Lagathu C, Sethi JK, Vidal-Puig A. Adipogenesis and WNT signalling. Trends Endocrinol Metab. 2009;20:16–24.

Jang MK, Kim CH, Seong JK, Jung MH. ATF3 inhibits adipocyte differentiation of 3T3-L1 cells. Biochem Biophys Res Commun. 2012;421:38–43.

Cheng CF, Ku HC, Cheng JJ, Chao SW, Li HF, Lai PF, et al. Adipocyte browning and resistance to obesity in mice is induced by expression of ATF3. Commun Biol. 2019;2:389.

Matsumoto M, Kondo K, Shiraki T, Brydun A, Funayama R, Nakayama K, et al. Genomewide approaches for BACH1 target genes in mouse embryonic fibroblasts showed BACH1-Pparg pathway in adipogenesis. Genes Cells. 2016;21:553–67.

Burrell JA, Boudreau A, Stephens JM. Latest advances in STAT signaling and function in adipocytes. Clin Sci (Lond). 2020;134:629–39.

Lee T, Sung MK, Lee S, Yang W, Oh J, Kim JY, et al. Convolutional neural network model to predict causal risk factors that share complex regulatory features. Nucleic Acids Res. 2019;47: e146.

Gao L, Uzun Y, Gao P, He B, Ma X, Wang J, et al. Identifying noncoding risk variants using disease-relevant gene regulatory networks. Nat Commun. 2018;9:702.

Hu ZL, Park CA, Reecy JM. Building a livestock genetic and genomic information knowledgebase through integrative developments of animal QTLdb and CorrDB. Nucleic Acids Res. 2019;47:D701–10.

Aibar S, Gonzalez-Blas CB, Moerman T, Huynh-Thu VA, Imrichova H, Hulselmans G, et al. SCENIC: single-cell regulatory network inference and clustering. Nat Methods. 2017;14:1083–6.

Ahmad B, Serpell CJ, Fong IL, Wong EH. Molecular mechanisms of adipogenesis: the anti-adipogenic role of AMP-activated protein kinase. Front Mol Biosci. 2020;7:76.

Zhao P, Stephens JM. Identification of STAT target genes in adipocytes. JAKSTAT. 2013;2: e23092.

PubMed   PubMed Central   Google Scholar  

Rosen ED, Hsu CH, Wang X, Sakai S, Freeman MW, Gonzalez FJ, et al. C/EPBalpha induces adipogenesis through PPARgamma: a unified pathway. Genes Dev. 2002;16:22–6.

Moseti D, Regassa A, Kim WK. Molecular regulation of adipogenesis and potential anti-adipogenic bioactive molecules. Int J Mol Sci. 2016;17:124.

Tanaka T, Yoshida N, Kishimoto T, Akira S. Defective adipocyte differentiation in mice lacking the C/EBPbeta and/or C/EBPdelta gene. EMBO J. 1997;16:7432–43.

Gonzales AM, Orlando RA. Role of adipocyte-derived lipoprotein lipase in adipocyte hypertrophy. Nutr Metab (Lond). 2007;4:22.

Hauberg ME, Fullard JF, Zhu L, Cohain AT, Giambartolomei C, Misir R, et al. Differential activity of transcribed enhancers in the prefrontal cortex of 537 cases with schizophrenia and controls. Mol Psychiatry. 2019;24:1685–95.

Kim TK, Hemberg M, Gray JM, Costa AM, Bear DM, Wu J, et al. Widespread transcription at neuronal activity-regulated enhancers. Nature. 2010;465:182–7.

Kapusta A, Kronenberg Z, Lynch VJ, Zhuo X, Ramsay L, Bourque G, et al. Transposable elements are major contributors to the origin, diversification, and regulation of vertebrate long noncoding RNAs. PLoS Genet. 2013;9: e1003470.

Crespo-Piazuelo D, Criado-Mesas L, Revilla M, Castello A, Noguera JL, Fernandez AI, et al. Identification of strong candidate genes for backfat and intramuscular fatty acid composition in three crosses based on the Iberian pig. Sci Rep. 2020;10:13962.

Zhang C, Wang Z, Bruce H, Kemp RA, Charagu P, Miar Y, et al. Genome-wide association studies (GWAS) identify a QTL close to PRKAG3 affecting meat pH and colour in crossbred commercial pigs. BMC Genet. 2015;16:33.

Falker-Gieske C, Blaj I, Preuss S, Bennewitz J, Thaller G, Tetens J. GWAS for meat and carcass traits using imputed sequence level genotypes in pooled F2-designs in pigs. G3 (Bethesda). 2019;9:2823–34.

Piorkowska K, Zukowski K, Ropka-Molik K, Tyra M, Gurgul A. A comprehensive transcriptome analysis of skeletal muscles in two Polish pig breeds differing in fat and meat quality traits. Genet Mol Biol. 2018;41:125–36.

Liu H, Xi Y, Liu G, Zhao Y, Li J, Lei M. Comparative transcriptomic analysis of skeletal muscle tissue during prenatal stages in Tongcheng and Yorkshire pig using RNA-seq. Funct Integr Genomics. 2018;18:195–209.

Xing K, Liu H, Zhang F, Liu Y, Shi Y, Ding X, et al. Identification of key genes affecting porcine fat deposition based on co-expression network analysis of weighted genes. J Anim Sci Biotechnol. 2021;12:100.

Jin L, Jiang Z, Xia Y, Lou P, Chen L, Wang H, et al. Genome-wide DNA methylation changes in skeletal muscle between young and middle-aged pigs. BMC Genomics. 2014;15:653.

Xu Z, Wu J, Zhou J, Zhang Y, Qiao M, Sun H, et al. Integration of ATAC-seq and RNA-seq analysis identifies key genes affecting intramuscular fat content in pigs. Front Nutr. 2022;9:1016956.

Fang L, Liu S, Liu M, Kang X, Lin S, Li B, et al. Functional annotation of the cattle genome through systematic discovery and characterization of chromatin states and butyrate-induced variations. BMC Biol. 2019;17:68.

Xiang R, van den Berg I, MacLeod IM, Hayes BJ, Prowse-Wilkins CP, Wang M, et al. Quantifying the contribution of sequence variants with regulatory and evolutionary significance to 34 bovine complex traits. Proc Natl Acad Sci USA. 2019;116:19398–408.

Cao X, Cheng J, Huang Y, Lan X, Lei C, Chen H. Comparative enhancer map of cattle muscle genome annotated by ATAC-seq. Front Vet Sci. 2021;8: 782409.

Kogelman LJ, Pant SD, Fredholm M, Kadarmideen HN. Systems genetics of obesity in an F2 pig model by genome-wide association, genetic network, and pathway analyses. Front Genet. 2014;5:214.

Download references

Acknowledgements

We would like to express our sincere gratitude to Dr. Wenlong Ma, and Mr. Weigang Zheng for their invaluable contributions to this project through insightful discussions.

This work was supported by the National Natural Science Foundation of China (Grant No. 31902124 and GZR20023 to L.Y.), the Shenzhen Science and Technology Innovation Commission (Grant No. JCYJ20190814163803664 to L.Y.), and the China National Key R&D Program during the 14th Five-year Plan Period (Grant No. 2021YFF1200503 and Grant No. 2021YFF1000600 to L.Y.).

Author information

Chao Wang and Choulin Chen contributed equally to this work.

Authors and Affiliations

Shenzhen Branch, Guangdong Laboratory for Lingnan Modern Agriculture, Key Laboratory of Livestock and Poultry Multi-Omics of MARA, Agricultural Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences, Shenzhen, 518124, People’s Republic of China

Chao Wang, Choulin Chen, Bowen Lei, Shenghua Qin, Yuanyuan Zhang, Kui Li, Song Zhang & Yuwen Liu

Innovation Group of Pig Genome Design and Breeding, Research Centre for Animal Genome, Agricultural Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences, Shenzhen, 518124, People’s Republic of China

Key Laboratory of Agricultural Animal Genetics, Breeding and Reproduction of Ministry of Education & Key Lab of Swine Genetics and Breeding of Ministry of Agriculture and Rural Affairs, Huazhong Agricultural University, Wuhan, 430070, People’s Republic of China

Chao Wang, Choulin Chen, Bowen Lei & Yuwen Liu

School of Life Sciences, Henan University, Kaifeng, 475004, People’s Republic of China

Yuanyuan Zhang

Shenzhen Research Institute of Henan University, Shenzhen, 518000, People’s Republic of China

Kunpeng Institute of Modern Agriculture at Foshan, Chinese Academy of Agricultural Sciences, Foshan, 528226, People’s Republic of China

You can also search for this author in PubMed   Google Scholar

Contributions

CW and YL conceived and designed the project. CW and BL analyzed the data. CC, SQ and YZ performed experiments. CW wrote the original draft. YL, KL and SZ reviewed and edited the manuscript. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Song Zhang or Yuwen Liu .

Ethics declarations

Ethics approval and consent to participate.

Not applicable.

Consent for publication

Competing interests.

The authors declare that they have no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1..

Simulating tissue-specific eRNA elements in the genome 1000 times.

Additional file 2: Table S1.

The GWAS-associated loci for muscle and adipose tissue-related traits in pigs.

Additional file 3: Table S2.

The list of primers for enhancer elements used in the self-transcribing active regulatory region sequencing (STARR-seq) system.

Additional file 4: Figure S1.

Detecting SNP regulatory activity in the self-transcribing active regulatory region sequencing (STARR-seq) system. ( a ) The library correlation in STARR-seq through a correlation analysis of count data in 22 eRNA regions. ( b ) SNP screening protocol for identifying regulatory activity in the STARR-seq system.

Additional file 5: Table S3.

The siRNA target sequence and RT-qPCR primers.

Additional file 6: Table S4.

Identified locations of enhancers in each tissue.

Additional file 7: Figure S2.

Identification methods and functional enrichment analysis of eRNAs. ( a ) Schematic diagram of unidirectional and bidirectional eRNA identification. ( b ) Gene ontology (GO) analysis reveals differential biological process pathways associated with eRNAs with distinct transcriptional directions. GO enrichment analysis was performed based on neighboring genes of eRNAs.

Additional file 8: Table S5.

Location information of detectable eRNAs in each tissue.

Additional file 9: Figure S3.

A heatmap revealing the dynamic expression atlas of detectable eRNAs in muscle and adipose tissues across breeds.

Additional file 10: Figure S4.

Pearson correlation between eRNA expression level and enhancer activity in specific tissue.

Additional file 11: Figure S5.

Ranked distribution plot of H3K27ac signal density identifies a small subset of super-enhancers. ( a ) Ranked distribution plot of H3K27ac signal density in Duroc fat, Duroc muscle ( b ), ES fat ( c ) and ES muscle ( d ) tissues, along with the number of identified super-enhancers.

Additional file 12: Table S6.

The expression activity of detectable eRNAs in different tissues.

Additional file 13: Table S7.

Muscle and fat tissue-specific expression of eRNAs and genes in eastern and western pigs.

Additional file 14: Figure S6.

Gene ontology (GO) analysis reveals biological process pathways relevant to non-tissue-specific eRNA expression. GO enrichment analysis was performed based on the neighboring genes of eRNAs.

Additional file 15: Figure S7.

Gene ontology (GO) analysis reveals biological process pathways relevant to genes expressed in a tissue- and breed-specific manner.

Additional file 16: Table S8.

The genome-wide association study (GWAS) enrichment results of tissue-specific eRNAs in pigs for homologous sequences in humans.

Additional file 17: Figure S8.

Characteristics and functional enrichment analysis of eRNA target genes in eRNA-mediated gene regulatory networks (eGRN). ( a ) Total number of eRNAs regulating various numbers of target genes. ( b ) Gene ontology (GO) analysis reveals biological functions of target genes regulated by tissue-specific expressed eRNAs in Duroc muscle and Enshi Black (ES) pig fat tissues. ( c ) Expression differences of target genes in muscle-related eGRN between eastern and western pigs (ES vs Duroc).

Additional file 18: Table S9.

eRNA target genes in eRNA-mediated gene regulatory networks (eGRN) related to muscle and adipose tissue.

Additional file 19: Table S10.

The list of selected SNPs for STARR-seq experiments.

Additional file 20: Table S11.

STARR-seq results for SNPs with significant regulatory activity.

Additional file 21: Table S12.

Regulatory relationships in fat-related hub eRNA target genes in eRNA-mediated gene regulatory networks (eGRN).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Wang, C., Chen, C., Lei, B. et al. Constructing eRNA-mediated gene regulatory networks to explore the genetic basis of muscle and fat-relevant traits in pigs. Genet Sel Evol 56 , 28 (2024). https://doi.org/10.1186/s12711-024-00897-4

Download citation

Received : 31 August 2023

Accepted : 03 April 2024

Published : 09 April 2024

DOI : https://doi.org/10.1186/s12711-024-00897-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Genetics Selection Evolution

ISSN: 1297-9686

• Submission enquiries: [email protected]
• General enquiries: [email protected]

Communication Research Strategies

• Comm Research Strategies
• The research process
• Defining your topic and crafting your research question
• Identifying search terms from your question
• Broaden or narrow your search
• Research methods
• Find background information
• Find journalism data and polls
• Find journal articles
• Find books and ebooks
• Find U.S. news articles
• Find international news articles
• Interlibrary loan
• How to distinguish between types of journal articles
• Components of a scholarly article, and things to consider when reading one
• Critically evaluating articles & other resources
• Literature reviews (opens a new guide) This link opens in a new window
• Annotated bibliographies
• Writing tools
• Citing sources (opens a new guide) This link opens in a new window
• Associated Press Style Manual
• Understanding & Avoiding Plagiarism (opens a new guide) This link opens in a new window

Contact me for research assistance

Components of a scholarly article

• Experimentation
• Results, analysis, & discussion
• Limitations
• Conclusion(s)

“What is this article about?”

Abstract : Most articles start with a paragraph called the “abstract”, which very briefly summarizes the whole article.

Introduction : This section introduces the research question under consideration in the article, and discusses what the article contributes to existing knowledge within the field.

Modified from the Department of Sociology, SUNY Brockport: https://www.brockport.edu/academics/sociology/journal.html

“What do we already know about this topic and what is left to discover?”

Literature review : the authors will review the existing research and theory on the topic, either as part of the introduction, or after the introduction under its own subtitle. The review of literature is meant to discuss previous work on the topic, point out what questions remain, and relate the research presented in the rest of the article to the existing literature.

“How did the author do the research?  Why this way and not this other way?”

Methodology : There is always some discussion of the methods used by the authors to conduct the study being reported.  If these is any sort of control group being used as part of the methodology, it will be described here.

"What specifically did the authors do?"

Experimentation : A description of the experimentation in which the authors engaged.  This section will report any data produced by the experiment, be it numerical data, qualitative response data, etc.  Data comes in many different forms.

“What did the authors find?  What did their findings reveal or say in response to their research question?”

Results, analysis, & discussion : A section or multiple sections will be devoted to analyzing the experiment that was conducted, as well as its results.  These sections then discuss the results, and what the analysis of these results revealed.

"What limitations does this research have, and how might they affect the conclusions that may be drawn from this research?"

Limitations: A description of any limitations of the experimentation in which the authors engaged, and how these limitations may affect the conclusions that may be drawn from this experimentation.

“What does it all mean and why is it important?”

Discussion and Conclusion : Articles typically end by discussing what the results mean and how the study contributes to existing knowledge. Answers to research questions are addressed in this section, as is whether any hypotheses were supported or not. The conclusion typically places the research in a larger context and discusses where future research on the topic may be headed.

Things to consider when reading a scholarly article / Shortcuts to reading journal articles

• Things to consider when reading a scholarly article
• Shortcuts to reading journal articles

Owen Lin. (2012, May 23). Deep Thoughts. Retrieved from https://www.flickr.com/photos/owen-pics/8406656015/. Used under the Creative Commons License.

• Know your research question or argument.  Though your question/argument may change or evolve as you delve deeper into the research process, you will want to have a solid idea of your research focus.  
• You don't have to read the entire article in order.  Start with the abstract which will give you a general summary of the article. If the abstract seems relevant then move to the conclusion or discussion section of the article to gain a better understanding of the article's main claims. At this point if the article does not seem relevant or useful then discard it. However, if the article does seem useful then spend as much time as necessary reading the article.  
• Read critically.  What is the author's argument? You will need to use your judgment when evaluating each source of information. Further research may be necessary if you find the author to be biased or you do not believe the validity of their argument.  
• Read the reference section.  Reading the references or works cited may lead you to other useful resources. You might also get a better understanding of the major players in the area you are researching.  
• Take notes.  How you do this is up to you. Make sure you keep your research question and argument in mind so you can be more efficient when taking notes. Created by Rachel Arteaga, for CSU Chico: http://libguides.csuchico.edu/c.php?g=462359&p=3163509

• Read the abstract first : Titles don’t always give much information. The abstract should give you just enough information to let you know the basics of the article. From this you will know whether you should read on or look elsewhere for your project. Some journals print a list of keywords pertaining to the article as well. These are further clues about the article.  
• Read the introduction and discussion/conclusion next : These sections will give you the main argument of the article, which should be helpful in determining its relevance to you and your project. You’ll also get a glimpse of the findings of the research being reported.  
• Read about the methods next : If what you’ve read so far interests you, get a sense of how the research was done. Is it a qualitative or quantitative project? What data are the study based on?  
• Read the Analysis and results next : If you decide that you are committed to this article, you should read in more detail about this research.

Created by the Department of Sociology, SUNY Brockport: https://www.brockport.edu/academics/sociology/journal.html

What makes information peer-reviewed vs. scholarly vs. non-scholarly? Which type of source should I use?

• What makes information peer-reviewed vs. scholarly vs. non-scholarly?
• Which type of source should I use?

There is a nuanced distinction between peer-review and scholarship, which typically doesn't matter when evaluating sources for possible citation in your own work.  Peer-review is a process through which editors of a journal have other experts in the field evaluate articles submitted to the journal for possible publication.  Different journals have different ways of defining an expert in the field.  Scholarly works, by contrast have an editorial process, but this process does not involve expert peer-reviewers.  Rather, one or more editors, who are themselves often highly decorated scholars in a field, evaluate submissions for possible publication.  This editorial process can be more economically driven than a peer-review process, with a greater emphasis on marketing and selling the published material, but as a general rule this distinction is trivial with regard to evaluating information for possible citation in your own work.

What is perhaps a more salient way of thinking about the peer-review / scholarship distinction is to recognize that while peer-reviewed information is typically highly authoritative, and is generally considered "good" information, the absence of a peer-review process doesn't automatically make information "bad."  More specifically, the only thing the absence of a peer-review process means is that information published in this manner is not peer-reviewed.  Nothing more.  Information that falls into this category is sometimes referred to as "non-scholarly" information -- but again, that doesn't mean this information is somehow necessarily problematic.

Where does that leave you in terms of deciding what type of information to use in producing your own work?  That is a highly individual decision that you must make.  The Which type of source should I use?  tab in this box offers further guidance on answering this question, though it is important to be aware that many WSU instructors will only consider peer-reviewed sources to be acceptable in the coursework you turn in .  You can ask your instructor for his or her thoughts on the types of sources s/he will accept in student work.

Image:  Martin Grater. (2017, Nov. 1). Deep Thought. Retrieved from https://www.flickr.com/photos/152721954@N05/24304490568/. Used under the Creative Commons License.

Your topic and research question or thesis statement will guide you on which resources are best.  Sources can be defined as primary, secondary and tertiary levels away from an event or original idea. Researchers may want to start with tertiary or secondary source for background information. Learning more about a topic will help most researchers make better use of primary sources.

While articles from scholarly journals are often the most prominent of the sources you will consider incorporating into your coursework, they are not the only sources available to you.  Which sources are most appropriate to your research is a direct consequence of they type of research question you decide to address.  In other words, while most university-level papers will require you to reference scholarly sources, not all will.  A student in an English course writing a paper analyzing Bob Dylan's lyrics, for example, may find an interview with Dylan published in Rolling Stone magazine a useful source to cite alongside other scholarly works of literary criticism.

The WSU University Libraries' What Sources Should I Use? handout, as well as the other sub-tabs under the  Evaluating information  section of this guide (which is indeed the section you are currently viewing) offer further guidance on understanding and identifying scholarly resources, and comparing them against different criteria to evaluate if they will be of value to your research.  How many non-scholarly works (if any) you are at liberty to cite alongside scholarly ones is often a question to ask of your professor.  Some may not want you to cite any, whereas others may be ok with some non-scholarly works cited alongside scholarly ones.

Image:  Brett Woods. (2006, Jan. 6). Deep Thoughts. Retrieved from https://www.flickr.com/photos/brettanicus/87653641/. Used under the Creative Commons License.

• << Previous: How to distinguish between types of journal articles
• Next: Critically evaluating articles & other resources >>
• Last Updated: Mar 27, 2024 2:38 PM
• URL: https://libraries.wichita.edu/comm-research-strategies

Advancing respirable coal mine dust source apportionment: a preliminary laboratory exploration of optical microscopy as a novel monitoring tool

• Open access
• Published: 16 April 2024
• Volume 11 , article number  30 , ( 2024 )

Cite this article

You have full access to this open access article

• Nestor Santa 1   na1 &
• Emily Sarver   ORCID: orcid.org/0000-0003-2301-8740 1   na1  

Exposure to respirable coal mine dust (RCMD) can cause chronic and debilitating lung diseases. Real-time monitoring capabilities are sought which can enable a better understanding of dust components and sources. In many underground mines, RCMD includes three primary components which can be loosely associated with three major dust sources: coal dust from the coal seam itself, silicates from the surrounding rock strata, and carbonates from the inert ‘rock dust’ products that are applied to mitigate explosion hazards. A monitor which can reliably partition RCMD between these three components could thus allow source apportionment. And tracking silicates, specifically, could be valuable since the most serious health risks are typically associated with this component-particularly if abundant in crystalline silica. Envisioning a monitoring concept based on field microscopy, and following up on prior research using polarized light, the aim of the current study was to build and test a model to classify respirable-sized particles as either coal, silicates, or carbonates. For model development, composite dust samples were generated in the laboratory by successively depositing dust from high-purity materials onto a sticky transparent substrate, and imaging after each deposition event such that the identity of each particle was known a priori . Model testing followed a similar approach, except that real geologic materials were used as the source for each dust component. Results showed that the model had an overall accuracy of \(86.5\%\) , indicating that a field-microscopy based monitor could support RCMD source apportionment and silicates tracking in some coal mines.

Avoid common mistakes on your manuscript.

1 Introduction

Exposure to respirable coal mine dust (RCDM) can cause lung diseases such as coal workers’ pneumoconiosis, progressive massive fibrosis, and chronic obstructive pulmonary disease (Laney and Attfield 2010 ; Hall et al. 2019 ; Liu et al. 2021 ; Almberg et al. 2023 ). A recent systematic review conducted by Shekarian et al. ( 2023 ) found that the resurgence of CWP among coal workers can be attributed to various factors, such as geographic location, mine size, mining operation type, coal-seam thickness, coal rank, changes in mining practices, technology advancement, and engineering dust control practices. Ongoing or resurgent prevalence of occupational lung disease among coal miners in the US and elsewhere has underscored the importance of effective RCMD monitoring strategies (National and of Sciences, Engineering, and Medicine. 2018 ). Acknowledging the role of crystalline silica in many cases, a 2018 consensus report published by the National Academies of Sciences, Engineering, and Medicine called for development of real-time silica monitoring capabilities; it also called for an overall better understanding of dust sources, and development and/or application of technologies to enable broad monitoring of RCMD trends (National and of Sciences, Engineering, and Medicine. 2018 ).

At present, real-time RCMD monitoring technologies only measure mass and particle concentrations, but not individual constituents. Indeed, there are only two types of devices currently certified as intrinsically safe for use in US coal mines: The continuous personal dust monitor (CPDM, such as the PDM3700; Thermo Fisher Scientific, Waltham, Massachusetts) is mandated for regulatory compliance monitoring (per 30 CFR part 74), and measures mass concentration of respirable dust using a tapered element oscillating microbalance (Patashnick and Rupprecht 1991 ; Colinet 2022 ). A device called the personal DataRAM (pDR-1000; Thermo Fisher Scientific, Waltham, Massachusetts) is also permissible-though it has been recently discontinued by the manufacturer-and measures particle concentration in the respirable range using a light scattering technique (Listak et al. 2007 ). Despite significant interest, real-time monitoring of specific RCMD constituents like crystalline silica has not yet come to fruition. Beyond the considerable costs of research and development, and the niche context of coal mining environments, there is also the challenge of analytical sensitivity. For example, silica might only represent a small percentage of the total RCMD mass, which itself is relatively small. However, a viable alternative to monitoring specific constituents of RCMD could be to track larger components, which are often correlated. Current methods for obtaining real-time data and detailed information about the components of dust particles are either lacking or insufficient. The need for monitoring capabilities that can quickly and comprehensively analyze various physical and chemical characteristics of RCMD, such as their size, shape, distribution, and chemical constituents, is crucial for the advancement of portable monitoring technologies (Gozzi et al. 2016 ).

In many coal mines, the primary components of dust can be loosely associated with three main sources: (1) coal is sourced from the target coal seam itself; (2) silicate minerals such as aluminosilicates and silica are frequently sourced from the rock strata surrounding the coal seam; and (3) carbonate minerals such as calcite are frequently sourced from inert ‘rock dust’ products that are applied to mitigate explosion risks (Agioutanti et al. 2020 ; Jaramillo et al. 2022 ; Sarver et al. 2021 ; Keles et al. 2022 ). While other constituents can be present in RCMD, such as metal sulfides or oxides, they typically do not account for a substantial portion of mass or particle count (Sarver et al. 2019 , 2021 ). Thus, in many mines, a simple partitioning of the RCMD into coal, silicates, and carbonates components in real-time (or near real-time) could enable source apportionment. Further, in mines where correlations can be established and periodically validated (e.g., via conventional sampling and laboratory analysis), tracking the silicates component of RCMD could serve as a proxy for silica.

Source apportionment is a well-established tool for understanding the origins of airborne particles and interpreting changes in composition (Thunis et al. 2019 ; Xue et al. 2022 ; Mondal and Singh 2021 ; Luo et al. 2022 ; Liu et al. 2022 ; Tohidi et al. 2022 ; Adeyemi et al. 2021 ; Das et al. 2020 ). Most applications in the mining context have been focused on understanding the contributions of diesel engines (Cantrell 1987 ; McDonald et al. 2003 ; Bugarski et al. 2020 ). With respect to RCMD, there have been several studies aimed at determining the sources of silica, specifically (Schatzel 2009 ; Keles et al. 2022 ), and establishing the relative source contributions of dust, more broadly (Jaramillo et al. 2022 ; Trechera et al. 2022 ). However, all of these efforts have relied on samples collected at discrete times and analyzed in the laboratory. A real-time capability for source apportionment could enable a better understanding of changes in RCMD with mining and geologic conditions or dust controls.

Envisioning a field microscopy based monitoring solution, prior research by the authors showed that respirable-sized dust particles could be classified as either coal or minerals using just a pair of images (Santa et al. 2021 ). To explain, after collecting dust particles onto a glass sampling substrate, static images were collected in both plane-polarized (PP) and cross-polarized (CP) light. The PP image was used to identify all particles in the image frame; and the CP image was used to identify mineral particles in the frame, since they typically illuminate based on their birefringence. Supposing samples can be collected and imaged on a semi-continuous basis, this approach could support a simple binary classification of RCMD (i.e., mineral versus coal). This could be valuable on its own in certain applications, like tracking the relative dust generation from mining in rock strata versus the target coal seam at the production face (Santa et al. 2021 ). Moreover, in an environment where silica content is understood to correlate well with the overall mineral component of RCMD, even a crude measurement of the mineral component might be valuable if made frequently.

Nevertheless, an obvious limitation of a binary mineral/coal classification scheme is that it does not distinguish between different mineral components, which might have different sources and/or relatively different associations with RCMD constituents of interest, such as silica. To support source apportionment, the total minerals fraction of the dust needs to be further partitioned. The earlier work by the authors hinted at the possibility of minerals subclassification, highlighting preliminary efforts to distinguish carbonates from silicates-which might be a better proxy than total minerals for tracking silica. However, particle size was identified as a constraining factor. It can impact classification accuracy, as smaller particles may prove more challenging to identify due to the limitations of microscope resolution and interference from other particles.

This preliminary laboratory-scale study aims to evaluate the technical feasibility of utilizing a field microscopy-based RCMD monitor for source apportionment in a limited number of samples and mine materials. The main objective is to develop and test a particle classification model using a direct measurement method that can effectively distinguish respirable-sized particles as coal, silicates, or carbonates. The study particularly examines the impact of particle size limits, as measured by projected area diameter (PAD), on improving the accuracy of the model.

2 Materials and methods

2.1 dust materials.

Four materials were used to develop the classification model: MIN-U-SIL ® 5 and MIN-U-SIL ® 10 (US Silica, Katy, TX, USA), which are high-purity quartz powders, were used as the source of silica particles. High-purity kaolinite powder (Ward’s Science, Rochester, NY, USA) was used as a representative silicate source. Clean bituminous coal and a real rock dust product (obtained from an industry partner) were used as representative sources of coal and carbonates, respectively. While the silica, kaolinite and rock dust were obtained as powders, the coal required milling to enable respirable-sized particles to be sampled. It was pulverized and sieved to obtain -230 mesh material ( \(< 63\,\upmu \textrm{m}\) ) as the source of coal dust.

For model testing, the same coal and rock dust were again used as dust sources. However, to better represent the range of silicate minerals that might occur in RCMD (i.e., beyond pure silica and kaolinite) and determine if the optical characteristics of the silicates in real materials were comparable, three real rock strata materials were obtained from industry partners. The two materials designated as “ROM rock” represent the rock strata that was mined at the production face in two different operations (Mine 11 and Mine 14); these were hand-sorted from the run-of-mine (ROM) material on the production belt and were pulverized and sieved prior to sampling of respirable-sized particles. The material designated as “bolter dust” was obtained directly from the dust collection system on a roof bolter machine at Mine 16. This material was already fine and required no preparation prior to respirable sampling.

To validate the purity of the seven materials used in this work, respirable-sized particles of each material were analyzed by scanning electron microscopy with energy dispersive X-ray (SEM-EDX). Briefly, a sample of each material was collected in the lab using a 10-mm nylon cyclone at \(2.0\,\textrm{L}/\textrm{min}\) to discard over-sized particles. The dust was collected on 37-mm polycarbonate (PC) filters in closed styrene cassettes. A 9-mm subsection of each filter was cut and prepared for analysis by sputter coating with Au/Pd. Then, the computer-controlled SEM-EDX routine described by Sarver et al. ( 2021 ), Johann-Essex et al. ( 2017 ) was used to identify, size and collect elemental data on about 500 particles per sample in the 1–10 \(\upmu \textrm{m}\) range. Per Sarver et al. ( 2021 ), particles were classified using their elemental content, and the mass percentage in each class was estimated using particle dimensions and assumed values for specific gravity. The SEM-EDX work was conducted using an FEI Quanta 600 FEG environmental scanning electron microscope (ESEM) (Hillsboro, OR, USA). This microscope was equipped with a backscatter electron detector (BSD) and a Bruker Quantax 400 EDX spectroscope (Ewing, NJ, USA). The following parameters were used for SEM-EDX analysis routines: 1000x magnification, 12.5 working distance, 15 kV accelerating voltage, and a spot size of \(5.5\,\upmu \textrm{m}\) . The resulting mass distribution of particles is shown in Table 1 . The carbonaceous class is typically composed of coal particles; the silicates consist of silica, aluminosilicates, and other silicates; the carbonates are made up of calcium and magnesium carbonates; and the others class comprises heavy minerals such as Fe and Ti-rich minerals.

2.2 Respirable dust sampling and imaging

For both model development and testing, samples of respirable-sized dust particles were prepared in two stages: particles from a single source material were deposited on the sample substrate in the first stage, and then particles from a different material were deposited on the same substrate in the second stage. Samples were imaged after each stage, with images being captured on multiple areas (“frames”) for each sample. The image frames were constant between stages, meaning that for a given sample the exact areas that were imaged after deposition of the first particle type were imaged again after deposition of the second particle type (Fig. 1 ). This process yielded a total of 34 samples (and 175 image frames) for model development, and another nine samples (153 frames) for model testing. The breakdown of samples and corresponding image frames is shown in Table 2 . (For each two-material combination, the table also shows the number of particles sourced from each material type. These values were determined comparing the Stage 1 and Stage 2 images using a “particle tracking” approach as described below.)

Two-stage method used to prepare and image samples

The two-stage dust sampling and imaging procedure is illustrated in Fig.  1 . This experimental design was specifically developed to allow direct classification of individual particles via particle tracking (described below). The sampling substrate was a small ( \(20\,\textrm{mm} \times 20\,\textrm{mm}\) ) piece of “sticky glass”. It consisted of a glass coverslip (AmScope, Irvine, CA, USA) which was overlaid with double-sided clear acrylic tape (Maxwel Manufacturing, Hangzhou, China) to minimize dust particle loss during handling. Prior to the first stage of dust sampling, target imaging areas were marked on the tape surface using colored fine-tip markers such that the same areas could be visited after each sampling stage. For sampling, the sticky glass was mounted on top of a cellulose filter pad inside a closed 37 mm styrene cassette (Zefon International, Ocala, FL, USA). For each sampling event, the desired dust material was aerosolized in a small enclosure using pulses of compressed air. A small air pump was used to pull dust laden air from the enclosure, through a a 10-mm nylon cyclone and then through the cassette. At \(2.0\,\textrm{L}/\textrm{min}\) , the cyclone particle penetration approximates the ISO respirable convention (International Organization for Standardization 1995 ).

When considering a sampling airflow rate of \(2.0\,\textrm{L}/\textrm{min}\) , the efficiency curve of the 10-mm Dorr-Oliver nylon cyclone closely approximates the respirable dust criterion, as outlined by Tomb and Treaftis ( 1983 ), Page and Volkwein ( 2009 ). As per ISO standard 7708:1995, respirable dust refers to a fraction of inhaled airborne particles capable of penetrating beyond the terminal bronchioles into the gas-exchange region of the lungs. This fraction comprises particles smaller than 10 microns, with a mass median diameter of 4 microns. It should be noted that the sampling efficiency of the nylon cyclone may be influenced by factors such as orientation, wind speed, and sampling flow rate, as observed by Kar and Gautam ( 1995 ). Nonetheless, the particle size distribution obtained from analyzing the microscope imagery from both the development and testing datasets indicates that the nylon cyclone can effectively separate the respirable fraction, as illustrated in Fig.  2 .

Particle size distribution for bulk materials used in Table 2

After the first stage of sampling (i.e., deposition of the first particle type), the sticky glass substrate was carefully removed and mounted onto a house-made 3D-printed sample holder which was used to maintain an x-y reference point for the microscope stage. Then, under the microscope, the color-coded image frame marks were used to locate the target areas. For each area, an Olympus BX53M Polarizing Microscope and Stream Start 2.3 imaging software (Olympus, Center Valley, PA, USA) were used to capture and save a pair of images: the first image was captured in transmitted plane-polarized light (TPP) and the second image was captured in transmitted cross-polarized light (TCP). Table 3 summarizes all the image acquisition settings, and a more detailed explanation of the settings was previously reported by Santa et al. ( 2021 ). After imaging, the substrate was carefully moved back to the sampling cassette and the second stage of particle deposition was completed with a different material. Finally, the same image frame areas were revisited to capture a new pair of images-again in TPP and TCP. In this way, the first pair of images only contained particles from the first material sampled, and the second pair of images contained particles from both materials.

2.3 Particle identification, feature extraction, and tracking

To identify individual particles in contrast to the background, the same particle identification approach detailed in Santa et al. ( 2021 ) was followed. Briefly, a local adaptive thresholding algorithm was used to classify image pixels into two categories: Foreground (particles) or background (substrate). Footnote 1 After performing morphological operations to clean edges and group particle pixels together, the output is a binary mask containing pixel category information. The binary mask was used to extract features for each individual particle including x and y location, size, circularity, and gray-scale intensity. In addition, the particle projected area diameter (PAD) was computed. Our analysis was limited to particles with a minimum PAD of \(1.5\,\upmu \textrm{m}\) to reduce the influence of optical signals from particles beyond the resolution capabilities of the microscope. This measure was taken to minimize the impact of potential noise caused by small particles.

To assign a “true” identity (i.e., silica, kaolinite, coal, or rock dust) to each individual particle in each composite sample a priori , a particle tracking algorithm (see Algorithm 1 ) was developed and applied to each set of TPP images (i.e., captured on the same frame area of a following Stage 1, and then Stage 2 particle deposition). In essence, it works to determine whether a particle is present in both images (i.e., first particle type) or only in the Stage 2 image (i.e., second particle type), which enables direct classification based on its source material. Figure  3 illustrates the particle tracking approach using images for a particular frame area on a sample that was first loaded with silica dust and then kaolinite dust. Using the binary masks described above, the tracking algorithm compares the x and y coordinates and circularity Footnote 2 for each particle identified in the Stage 2 image to each particle identified in the Stage 1 image. If a Stage 2 image and Stage 1 image particle match, the particle is assigned the Stage 1 particle type (e.g., silica in the example shown in Fig.  3 ); if a Stage 2 image particle does not have a match in the Stage 1 image, it is assigned the Stage 2 particle type (e.g., kaolinite in the example shown in Fig.  3 ). A match is defined as when the Euclidean distance (Eq. ( 1 )) between particles in the feature space (i.e., for x and y coordinates, and circularity) is below some threshold. To determine the threshold for this task, a subset of the samples and image frames collected for model development were inspected to ensure the results matched the expected outcome closely. The labels obtained by the particle tracking algorithm underwent a thorough comparative analysis against stage 1 and stage 2 images, whereby each particle’s anticipated label was manually verified against its corresponding determined label in a subset of model development samples. The threshold value was chosen to maximize accuracy in particle label assignments. While the current work only included two stages of particle deposition and imaging, the same approach could theoretically be applied to samples generated using more stages.

Particle tracking approach

Algorithm 1. Algorithm used to identify and track dust particles collected during sequential deposition

Figure  4 shows additional examples of particle tracking results for sets of Stage 1 and Stage 2 TPP images captured on three different samples (i.e., with silica + coal, silica + kaolinite, or silica + rock dust). These examples clearly illustrate how the particle tracking approach can be used for direct classification of individual particles that are loaded in stages from high-purity source materials. Thus, while the classification model being developed is intended to classify particles in real, composite samples (i.e., something akin to the Stage 2 images alone), the particle tracking results (i.e., as shown in Table 2 ) served as the “true” classification for each particle during model development. This allowed direct evaluation of model accuracy.

Images of dust samples ( a – c ) initially loaded with silica during stage 1 under both TPP and TCP lighting conditions. Images in stage 2, show coal, kaolinite, and rock dust particles deposited on samples a – c , respectively. On TPP images, the particle boundaries have been determined for silica (orange), coal, kaolinite, and rock dust (blue) to show automated particle tracking between dust-loading events. Annotations (red) highlight dim silica particles on samples a – c during sequential dust deposition

3 Results and discussion

3.1 model development, 3.1.1 particle classification.

The images presented in Fig.  4 illustrate the distinctions in birefringence among various particle types. Notably, coal particles exhibit no birefringence, silica particles are characterized by minimal birefringence, kaolinite particles demonstrate moderate birefringence, while rock dust particles exhibit high birefringence. Such variations may be effectively utilized by a classification model in order to discern and categorize dust particles.

To build the particle classification model, all 175 pairs of TPP and TCP images that were captured following Stage 2 of dust particle deposition were used (i.e., see image frames for model development in Table 2 ). Based on the particle tracking algorithm, these included 1087 silica particles and 1769 kaolinite particles (2856 silicate particles in total), 1947 coal particles, and 508 rock dust particles. As stated, the TPP binary mask was used to extract features for each particle, including the grayscale intensity. To establish classification thresholds, two grayscale intensity metrics were used: (1) The “added mean particle intensity” (AMPI) was calculated by summing the mean particle intensity observed in a TPP image with that in the paired TCP image. The AMPI exploits the birefringence of particles under TCP light to detect differences between minerals and coal. In RCMD, most minerals are expected to have anisotropic crystalline structures and thus be birefringent to some extent, whereas the coal is expected to be non-crystalline and thus non-birefringent (The AMPI metric was described in the earlier work by Santa et al. ( 2021 )). (2) The “multiplication of the mean grayscale intensity” (MMPI) was calculated by multiplying the mean particle intensity observed in the TPP with that observed in the paired TCP image. The MMPI serves as a measure of the relative change in particle brightness between lighting conditions. It is expected that highly birefringent particles will have relatively higher MMPI values than less birefringent particles.

Figure  5 shows the distribution of AMPI and MMPI values, for all particles identified in the model development data set. For these figures, the results have been split into two categories to visualize the effect of particle size (as measured by PAD). The split between low and high PAD was made based on the median value observed across all particles in the model development data set ( n = 5311). Consistent with expectations based on the earlier work (Santa et al. 2021 ), higher PAD improved the separation between coal and mineral particles using the AMPI. Higher PAD also yielded improved separation between the mineral particle types (i.e., the two silicates versus rock dust) using MMPI. Given the relatively small number of particles with very high PAD included in the model development data set, additional splits in the data were not considered for the current work-though this should certainly be a focus of future efforts. Further, while reliable separation of silica and kaolinite using the MMPI does not appear feasible under the PAD conditions represented in Fig.  5 , this could also be a topic for future research that seeks to classify silica, specifically. It is noted that, while particle loading density, i.e., number of particles within the image frame, did not appear to be an issue in the current work, this is a factor that likely would need to be controlled in a practical application. Simply put, if particles are in close proximity, they might interfere with one another (Santa et al. 2021 ).

Using the high PAD subset of the model development data shown in Fig.  5 , AMPI and MMPI thresholds were set to separate the three primary components (i.e., coal, silicates and carbonates). From here, kaolinite and silica have been grouped together as ’silicates’, and rock dust is now ’carbonates’. The AMPI and MMPI thresholds shown in Fig.  6 form the basis of a two-step classification model: in the first step, AMPI is used to classify a particle as either coal or mineral, and in the second step, MMPI is used to subclassify mineral particles as either silicate or carbonate. The threshold values were determined to minimize the differences between precision and recall across all three classes.

Precision is a statistical benchmark used to assess the accuracy of a model’s positive predictions. To determine the precision, the number of true positive instances is divided by the total number of positive predictions made by the model Eq. ( 2 ). This proportion essentially quantifies the ability to predict positive instances correctly. (A high precision score indicates there are few false positives in the model predictions.) Recall, on the other hand, evaluates a model’s ability to correctly identify all positive instances within a given dataset Eq. ( 3 ). Essentially, this measure gauges whether the model can capture all positive instances without missing any. Thus, recall is important for assessing false negatives (a high recall score indicates there are few false negatives.)

In the domain of classification problems, the metrics of precision and recall are often found to be in tension with one another, with an increase in precision leading to a reduction in recall, and vice versa. The choice of whether to optimize for precision or recall is dependent on the specific requirements of the problem. In this study, the two metrics were balanced to assess the overall performance of the model. To achieve this balance, the precision and recall values for each of the three classes (coal, silicates, carbonates) were evaluated, resulting in six values that were used to calculate the standard deviation. The standard deviation was used to determine the degree of difference between the values, with a high standard deviation indicating greater divergence and a low standard deviation indicating a more optimal balance.

A simple algorithm was employed to identify the combination of AMPI and MMPI thresholds that resulted in the minimum standard deviation (11.22%). The values of precision and recall that minimize the standard deviation are summarized in Fig.  6 c. Based on the selected thresholds, an overall accuracy of 86.5% was observed for particles included in the model development image inventory Table 2 having PAD \(> 2.47\,\upmu \textrm{m}\) . As shown in Fig.  6 , much of the misclassification was attributed to the overlap between the silicates and carbonates classes. In essence, some rock dust particles had relatively low MMPI values, while some silica and kaolinite particles had relatively high values.

Distribution of AMPI values observed for all particles in the model development data set. Data are split into two groups based on whether the PAD is above or below the median value observed for the entire data set ( \(2.47\,\upmu \textrm{m}\) )

Distributions of a  AMPI and b  MMPI showing optimal threshold values, with an overall accuracy of 86.5% c  The confusion matrix displays the overall count of observations in regard to true and predicted categories

3.2 Model testing

Next, the two-step classification model was challenged using the high PAD particles (PAD \(> 2.47\,\upmu \textrm{m}\) ) included in the testing dataset shown in Table 2 . As mentioned, the key difference between the dust particles in this data set and those used for model development was that the silicate particles were sourced from real mine materials (i.e., ROM rock and bolter dust, expected to contain a range of silicate minerals) rather than pure silica and kaolinite powders. It is noted that the real coal and rock dust particle sources were the same for both data sets, since these particle types are expected to be more similar across mines in terms of their optical features. I.e., Coal is generally expected to be non-birefringent, and carbonates are generally expected to be sourced from rock dust products which are often made from high purity calcite. The samples for model testing were generated and imaged using the same two-stage procedure used for the model development samples, and the images were processed using the same algorithms to find particles and extract feature data. Moreover, the particle tracking algorithm was applied to define the true source of each particle in the testing data set.

Figure  7 compares the classification model results (i.e., based on particle AMPI and MMPI observed in the Stage 2 composite sample images) to the results of the particle tracking algorithm (i.e., particles classified based on their source material). On average the percentage difference across all results was 4%, and across all particle classes the average error was 4.54% for coal, 5.12% for silicates, and 2.34% for carbonates. The highest percentage difference between the predicted class and true source was 15.4% for estimating the silicates fraction in the sample containing Mine 16 bolter dust and coal (Mine 16 BD+C). Some discrepancies are probably due to misclassifications, which expected to happen more often between silicates and carbonates per Fig.  6 . This is because there is some overlap between the range of MMPI values for silicate and carbonate particles. Additionally, silicate particles which have low birefringence may be incorrectly identified as coal particles (i.e., there is also some overlap in the range of AMPI values for silicate and coal particles).

Based on the higher average error noted above within the silicate class, it may be necessary to conduct calibration efforts on a mine-by-mine basis if this method were to be implemented in the field. This is important because if the silicates class holds for a broader range of minerals sourced from rock strata, such a classification scheme could help track the primary dust sources in coal mines (i.e., coal, rock dust, and rock strata).

Regarding the results shown in Fig.  7 , another factor in observed discrepancies between the model classification and particle tracking results could be due to impurities present in the real ROM and BD source material. Impurities would not accounted for in the particle tracking approach to direct identification, because the tracking approach attempts to label a particle with its true source. This should work well for particles sourced from highly pure materials, but real source materials have some impurities. For example, per Table 1 , respirable dust generated from the Mine 16 bolter dust material was dominated by silicates, but it had some carbonaceous particles. This might explain why the model predicts somewhat more coal than the particle tracking approach for samples containing this material. The consideration of the margin of error in real mine materials when utilizing the particle tracking approach is a crucial factor that can impact the accuracy of the model. Such errors can be carried over into the model-developing process and must be taken into account in any analysis of the data.

Nevertheless, the particle tracking approach developed for this work represents an important improvement upon the earlier work by the authors (Santa et al. 2021 ). Previously, images were not captured between sampling stages to identify particles in composite samples directly. Rather, SEM-EDX was employed for reference measurements. This was done by analyzing a paired sample for each composite sample that was imaged under the optical microscope, such that the overall class distribution determined by each method could be compared. Clearly, a more direct reference measurement (i.e., that does not require a separate analytical method on a separate sample) is superior for both efficiency and for minimizing uncertainties and errors.

Although source apportionment is the main objective of this study, the results presented here also imply that subclassification of silica particles might be possible in a particular PAD range. Specifically, particle classification may be improved as size increases. As noted above, when using a crude split between high and low PAD datasets, the results suggest that the classification of silica will be challenging. I.e., Even when considering particles with high PAD, there is a significant overlap between the silica and kaolinite MMPI values in Fig.  5 . However, as PAD increases, differences in birefringence start to emerge. These results beg for further research in a somewhat larger PAD range.

While somewhat larger particles are not “respirable” by definition, some previous findings indicate that the distribution of particles in perhaps the 10–20 \(\upmu \textrm{m}\) range might still be representative of the respirable distribution. For instance, analysis of airborne coal mine dust sampled with a respirable cyclone and with custom cyclones designed to cut at 10 or \(20\,\upmu \textrm{m}\) showed that dust composition does not change much between those three fractions (Animah et al. 2023 ). Thus, the silica content in the 10–20 \(\upmu \textrm{m}\) range may be a good proxy for silica in the respirable range-especially if field microscopy based measurements can be initially calibrated to another method and validated periodically in a given environment.

Analysis on testing samples containing respirable dust deposited from three mine materials during stage one (mine 11 ROM rock, mine 14 ROM rock, and mine 16 bolter dust), followed by source dust deposition during stage two (rock dust or coal). Darker colors represent the results of the AMPI-MMPI classification model shown in Fig.  6 . Lighter colors denote the results obtained by the particle tracking (PT) approach

4 Conclusions

The preliminary laboratory work presented here demonstrates that a simple optical microscopy approach could serve as the basis for field monitoring of RCMD. Particles in the respirable range can be separated into 3 main classes, which can support source apportionment in many mines. In addition, tracking the silica component in the larger fraction specifically could be a valuable proxy for silica in some mines.

If this method is to be used in coal mines, it may be necessary to conduct calibration and periodic validation work. This is crucial because proper classification of minerals can help identify the main sources of dust, including coal, rock dust, and rock strata.

The work here also implies the possibility for silica classification under favorable PAD conditions. The next step should focus on determining optimal particle size and loading density to subclassify silica particles.

The particle tracking approach utilized here is a useful technique for direct reference measurement, eliminating the need for SEM-EDX and increasing measurement efficiency while reducing uncertainty. Unlike SEM, the particle tracking method estimates dust constituents directly from image frames, allowing for high-confidence particle identification.

Finally, although samples have been prepared to contain the main dust constituents present in underground coal environments, and the materials used were all obtained from actual mine materials, further research should also include filter samples directly obtained from sampling campaigns in underground coal mines. This preliminary study is an important initial step towards comprehending the constituents and sources of RCMD. In order to achieve a comprehensive source apportionment analysis in the future, a wider range of dust materials from various mines will be required to better represent the complexity of RCMD in underground coal environments and assess the performance of the method described in this study.

Data availability

Not applicable

Only the TPP images were used for identifying particles as they were found to be the most effective for this task in earlier work (Santa et al. 2021 ).

x , y , and circularity were chosen as features for particle tracking working under the assumption that the particle location and shape should not dramatically change between sequential dust collection.

Adeyemi A, Molnar P, Boman J, Wichmann J (2021) Source apportionment of fine atmospheric particles using positive matrix factorization in Pretoria, South Africa. Environ Monit Assess 193:1573–2959. https://doi.org/10.1007/s10661-021-09483-3

Article   Google Scholar  

Agioutanti E, Keles C, Sarver E (2020) A thermogravimetric analysis application to determine coal, carbonate, and non-carbonate minerals mass fractions in respirable mine dust. J Occup Environ Hyg 17(2–3):47–58. https://doi.org/10.1080/15459624.2019.1695057

Almberg KS, Halldin CN, Friedman LS, Go LHT, Rose CS, Hall NB, Cohen RA (2023) Increased odds of mortality from non-malignant respiratory disease and lung cancer are highest among us coal miners born after 1939. Occup Environ Med 80(3):121–128. https://doi.org/10.1136/oemed-2022-108539

Animah F, Greth A, Keles C, Sarver E (2023) Effect of auxiliary scrubbers on respirable coal mine dust particle size and composition. In: Underground ventilation, pp 221–229

Bugarski AD, Hummer JA, Vanderslice S, Shahan MR (2020) Characterization of aerosols in an underground mine during a longwall move. Min Metall Explor 37(4):1065–1078. https://doi.org/10.1007/s42461-020-00209-6

Cantrell BK (1987) Source apportionment analysis applied to mine dust aerosols: coal dust and diesel emissions aerosol measurement. In: Proceedings of the 3rd mine ventilation symposium, pp 495–501

Colinet JF, Mischler SE (2022) Effectiveness of the cpdm in reducing overexposures to coal mine dust. Min Metall Explor 39:283–290. https://doi.org/10.1007/s42461-022-00563-7

Das A, Kumar R, Patel SS, Saha MC, Guha D (2020) Source apportionment of potentially toxic elements in street dust of a coal mining area in Chhattisgarh, India, using multivariate and lead isotopic ratio analysis. Environ Monit Assess 192:1573–2959. https://doi.org/10.1007/s10661-020-08321-2

Gozzi F, Della Ventura G, Marcelli A (2016) Mobile monitoring of particulate matter: state of art and perspectives. Atmos Pollut Res 7(2):228–234. https://doi.org/10.1016/j.apr.2015.09.007

Hall NB, Blackley DJ, Halldin CN, Laney AS (2019) Current review of pneumoconiosis among us coal miners. Curr Environ Health Rep 6:137–147. https://doi.org/10.1007/s40572-019-00237-5

International Organization for Standardization (ISO) (1995) Air quality: particle size fraction definitions for health-related sampling. Switzerland, Geneva

Jaramillo L, Agioutanti E, Ghaychi Afrouz S, Keles C, Sarver E (2022) Thermogravimetric analysis of respirable coal mine dust for simple source apportionment. J Occup Environ Hyg 19(9):568–579. https://doi.org/10.1080/15459624.2022.2100409

Johann-Essex V, Keles C, Sarver E (2017) A computer-controlled sem-edx routine for characterizing respirable coal mine dust. Minerals 7(1):15. https://doi.org/10.3390/min7010015

Kar K, Gautam M (1995) Orientation bias of the isolated 10-mm nylon cyclone at low stream velocity. Am Ind Hyg Assoc J 56(11):1090–1098. https://doi.org/10.1080/15428119591016304

Keles C, Pokhrel N, Sarver E (2022) A study of respirable silica in underground coal mines: Sources. Minerals 12(9):1115. https://doi.org/10.3390/min12091115

Laney AS, Attfield MD (2010) Coal workers’ pneumoconiosis and progressive massive fibrosis are increasingly more prevalent among workers in small underground coal mines in the united states. Occup Environ Med 67(6):428–431. https://doi.org/10.1136/oem.2009.050757

Listak JM, Chekan GJ, Colinet JF, Rider JP (2007) Performance of a light scattering dust monitor at various air velocities: results of sampling in the active versus the passive mode. Int J Min Resources Eng 12(1):35–47

Google Scholar  

Liu S, Cheng W, Wang G, Fan L, Zhang R (2021) 4. Special issue on mine dust research: health effects and control technologies. Int J Coal Sci Technol 8(2):177–178. https://doi.org/10.1007/s40789-021-00434-4

Liu Y, Wang R, Zhao T, Zhang Y, Wang J, Wu H, Hu P (2022) Source apportionment and health risk due to pm10 and tsp at the surface workings of an underground coal mine in the arid desert region of northwestern china. Sci Total Environ 803:149901. https://doi.org/10.1016/j.scitotenv.2021.149901

Luo H, Wang Q, Guan Q, Ma Y, Ni F, Yang E, Zhang J (2022) Heavy metal pollution levels, source apportionment and risk assessment in dust storms in key cities in northwest china. J Hazard Mater 422:126878. https://doi.org/10.1016/j.jhazmat.2021.126878

McDonald JD, Zielinska B, Sagebiel JC, McDaniel MR, Mousset-Jones P (2003) Source apportionment of airborne fine particulate matter in an underground mine. J Air Waste Manag Assoc 53(4):386–395. https://doi.org/10.1080/10473289.2003.10466178

Mondal S, Singh G (2021) Pm2.5-bound trace elements in a critically polluted industrial coal belt of India: seasonal patterns, source identification, and human health risk assessment. Environ Sci Pollut Res 28(25):32634–32647. https://doi.org/10.1007/s11356-021-12876-z

National Academies, of Sciences, Engineering, and Medicine. (2018) Monitoring and Sampling Approaches to Assess Underground Coal Mine Dust Exposures. The National Academies Press, Washington, DC

Page SJ, Volkwein JC (2009) A revised conversion factor relating respirable dust concentrations measured by 10 mm dorr-oliver nylon cyclones operated at 1.7 and 2.0 l min-1. J Environ Monit 11:684–689. https://doi.org/10.1039/B817922K

Patashnick H, Rupprecht EG (1991) Continuous pm-10 measurements using the tapered element oscillating microbalance. J Air Waste Manag Assoc 41(8):1079–1083. https://doi.org/10.1080/10473289.1991.10466903

Santa N, Keles C, Saylor JR, Sarver E (2021) Demonstration of optical microscopy and image processing to classify respirable coal mine dust particles. Minerals 11(8):838. https://doi.org/10.3390/min11080838

Sarver E, Keles C, Afrouz SG (2021) Particle size and mineralogy distributions in respirable dust samples from 25 us underground coal mines. Int J Coal Geol 247:103851. https://doi.org/10.1016/j.coal.2021.103851

Sarver E, Keles C, Rezaee M (2019) Beyond conventional metrics: comprehensive characterization of respirable coal mine dust. Int J Coal Geol 207:84–95. https://doi.org/10.1016/j.coal.2019.03.015

Schatzel SJ (2009) Identifying sources of respirable quartz and silica dust in underground coal mines in southern west Virginia, western Virginia, and eastern kentucky. Int J Coal Geol 78(2):110–118. https://doi.org/10.1016/j.coal.2009.01.003

Shekarian Y, Rahimi E, Rezaee M, Roghanchi P (2023) 5. A systematic review of occupational exposure to respirable coal mine dust (RCMD) in the U.S. mining industry. Int J Coal Sci Technol 10(1):29. https://doi.org/10.1007/s40789-023-00586-5

Thunis P, Clappier A, Tarrason L, Cuvelier C, Monteiro A, Pisoni E, Wesseling J, Belis CA, Pirovano G, Janssen S, Guerreiro C, Peduzzi E (2019) Source apportionment to support air quality planning: strengths and weaknesses of existing approaches. Environ Int 130:104825. https://doi.org/10.1016/j.envint.2019.05.019

Tohidi R, Farahani VJ, Sioutas C (2022) Real-time measurements of mineral dust concentration in coarse particulate matter ( \({PM_{10-2.5}}\) ) by employing a novel optical-based technique in los angeles. Sci Total Environ 838:156215. https://doi.org/10.1016/j.scitotenv.2022.156215

Tomb TF, Treaftis HN (1983) 6. Review of published de83 902993 experimental calibrations performed on the 10-millimeter nylon cyclone. Technical report, Mining Enforcement and Safety Administration

Trechera P, Querol X, Lah R, Johnson D, Wrana A, Williamson B, Moreno T (2022) Chemistry and particle size distribution of respirable coal dust in underground mines in central eastern Europe. Int J Coal Sci Technol 9(1):3. https://doi.org/10.1007/s40789-022-00468-2

Xue F, Niu H, Hu S, Wu C, Zhang C, Gao N, Ren X, Li S, Hu W, Wang J, Fan J (2022) Seasonal variations and source apportionment of carbonaceous aerosol in pm2.5 from a coal mining city in the north china plain. Energy Explor Exploit 40(2):834–851. https://doi.org/10.1177/01445987211043211

Download references

Acknowledgements

The authors gratefully acknowledge the Alpha Foundation for the Improvement of Mine Safety and Health for funding this work. We also thank our industry partners for providing the dust materials used in this study. The opinions and viewpoints expressed in this work are solely those of the authors, and do not necessarily reflect those of our research sponsors or partners.

This work was supported by the Alpha Foundation for the Improvement of Mine Safety and Health, grant number AFC316FO-84.

Author information

Nestor Santa and Emily Sarver have contributed equally to this work.

Authors and Affiliations

Department of Mining and Minerals Engineering, Virginia Tech, Blacksburg, VA, 24061, USA

Nestor Santa & Emily Sarver

You can also search for this author in PubMed   Google Scholar

Contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by NS. The first draft of the manuscript was written by NS and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Emily Sarver .

Ethics declarations

Ethical approval and consent to participate, consent for publication, competing interests.

The authors have no competing interests to declare that are relevant to the content of this article.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Santa, N., Sarver, E. Advancing respirable coal mine dust source apportionment: a preliminary laboratory exploration of optical microscopy as a novel monitoring tool. Int J Coal Sci Technol 11 , 30 (2024). https://doi.org/10.1007/s40789-024-00687-9

Download citation

Received : 19 October 2023

Revised : 26 November 2023

Accepted : 14 March 2024

Published : 16 April 2024

DOI : https://doi.org/10.1007/s40789-024-00687-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Polarized light microscopy
• Image processing
• Dust monitoring
• Respirable silica
• Coal mining
• Find a journal
• Publish with us
• Track your research

This paper is in the following e-collection/theme issue:

Published on 17.4.2024 in Vol 26 (2024)

Mobile Apps to Support Mental Health Response in Natural Disasters: Scoping Review

Authors of this article:

• Nwamaka Alexandra Ezeonu 1 , MBBS, MSc, MBA   ; 
• Attila J Hertelendy 2, 3 , BSc, MHS, MSc, PhD   ; 
• Medard Kofi Adu 4 , BSc, MSc   ; 
• Janice Y Kung 5 , BCom, LMIS   ; 
• Ijeoma Uchenna Itanyi 1, 6, 7 , MBBS, MPH   ; 
• Raquel da Luz Dias 4 , BSc, MSc, PhD   ; 
• Belinda Agyapong 8 , HDip, BSc, MEd   ; 
• Petra Hertelendy 9 , BS   ; 
• Francis Ohanyido 10 , MBBS, MBA, MPH   ; 
• Vincent Israel Opoku Agyapong 4 , BSc, PGD, MBChB, MSc, MD, PhD   ; 
• Ejemai Eboreime 4 , MBBS, MSc, PhD  

1 Center for Translation and Implementation Research, College of Medicine, University of Nigeria, Nsukka, Nigeria

2 Department of Information Systems and Business Analytics, College of Business, Florida International University, Miami, FL, United States

3 Department of Emergency Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, United States

4 Department of Psychiatry, Faculty of Medicine, Dalhousie University, Halifax, NS, Canada

5 Geoffrey and Robyn Sperber Health Sciences Library, University of Alberta, Edmonton, AB, Canada

6 Department of Community Medicine, University of Nigeria, Enugu, Nigeria

7 Department of Public Health Sciences, Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada

8 Department of Psychiatry, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB, Canada

9 Department of Psychology, Florida State University, Tallahassee, FL, United States

10 West African Institute of Public Health, Abuja, Nigeria

Corresponding Author:

Ejemai Eboreime, MBBS, MSc, PhD

Department of Psychiatry

Faculty of Medicine

Dalhousie University

5909 Veterans' Memorial Lane

8th Floor Abbie J Lane Memorial Building, QEII Health Sciences Centre

Halifax, NS, B3H 2E2

Phone: 1 9024732479

Email: [email protected]

Background: Disasters are becoming more frequent due to the impact of extreme weather events attributed to climate change, causing loss of lives, property, and psychological trauma. Mental health response to disasters emphasizes prevention and mitigation, and mobile health (mHealth) apps have been used for mental health promotion and treatment. However, little is known about their use in the mental health components of disaster management.

Objective: This scoping review was conducted to explore the use of mobile phone apps for mental health responses to natural disasters and to identify gaps in the literature.

Methods: We identified relevant keywords and subject headings and conducted comprehensive searches in 6 electronic databases. Studies in which participants were exposed to a man-made disaster were included if the sample also included some participants exposed to a natural hazard. Only full-text studies published in English were included. The initial titles and abstracts of the unique papers were screened by 2 independent review authors. Full texts of the selected papers that met the inclusion criteria were reviewed by the 2 independent reviewers. Data were extracted from each selected full-text paper and synthesized using a narrative approach based on the outcome measures, duration, frequency of use of the mobile phone apps, and the outcomes. This scoping review was reported according to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews).

Results: Of the 1398 papers retrieved, 5 were included in this review. A total of 3 studies were conducted on participants exposed to psychological stress following a disaster while 2 were for disaster relief workers. The mobile phone apps for the interventions included Training for Life Skills, Sonoma Rises, Headspace, Psychological First Aid, and Substance Abuse and Mental Health Services Administration (SAMHSA) Behavioural Health Disaster Response Apps. The different studies assessed the effectiveness or efficacy of the mobile app, feasibility, acceptability, and characteristics of app use or predictors of use. Different measures were used to assess the effectiveness of the apps’ use as either the primary or secondary outcome.

Conclusions: A limited number of studies are exploring the use of mobile phone apps for mental health responses to disasters. The 5 studies included in this review showed promising results. Mobile apps have the potential to provide effective mental health support before, during, and after disasters. However, further research is needed to explore the potential of mobile phone apps in mental health responses to all hazards.

Introduction

Rising global average temperatures and associated changes in weather patterns result in extreme weather events that include hazards such as heatwaves, wildfires, hurricanes, floods, and droughts [ 1 ]. These extreme events linked to climate change are resulting in overlapping and so-called cascading disasters leading to record numbers of “billion dollar” disasters with significant losses of lives and property [ 2 , 3 ]. In 2021 alone, approximately 10,000 fatalities caused by disasters were reported globally, while the economic loss was estimated at approximately US $343 billion [ 4 ]. Disasters are predicted to become more recurring as a result of the impact of human activities such as burning fossil fuels and deforestation, which release greenhouse gases into the atmosphere that trap heat and cause global temperatures to rise [ 5 ].

These catastrophes can adversely affect physical health, mental health, and well-being in both the short and long term as a result of changes due to the political and socioeconomic content, evacuations, social disruption, damage to health care facilities, and financial losses [ 6 - 10 ]. It is estimated that about 33% of people directly exposed to natural disasters will experience mental health sequelae such as posttraumatic stress disorders (PTSDs), anxiety, and depression, among others [ 11 , 12 ].

There is growing recognition of the importance of incorporating mental health into medical and emergency aspects of disaster response [ 12 , 13 ]. However, in contrast to most medical response strategies that are largely curative, mental health response to disasters is predicated on the principles of preventive medicine, thus, emphasizing health promotion, disaster prevention, preparedness, and mitigation [ 14 ]. The strategies of mental health response span across primary prevention (mitigating the risk of ill health before it develops), secondary prevention (early detection and intervention), and tertiary prevention (managing established ailment and averting further complications) [ 15 ].

Mobile health (mHealth) technology has shown great promise in mental health and has been applied across the 3 levels of prevention [ 16 - 20 ]. For example, SMS text messaging and mobile apps have been developed to promote mental health awareness among young people and older adults (primary prevention) [ 21 ]. Additionally, during the COVID-19 pandemic, mHealth was deployed at the population level in Canada to screen for symptoms of anxiety and depression (secondary prevention) [ 22 ]. In addition, mHealth interventions were deployed to support first responders and essential workers during the pandemic [ 23 , 24 ]. Further, the technology has been deployed for therapeutic purposes in patients diagnosed with mental health conditions while simultaneously providing support against complications such as suicidal ideation (tertiary prevention) [ 25 ].

Although videoconferencing and phone calls can be used for mental health conditions, mobile apps provide more mobility and accessibility, are interactive, more adaptable to users’ routines, and can be used repeatedly [ 26 , 27 ]. While numerous academic studies have been conducted on the app of mHealth in the preventive and curative management of mental health conditions in clinical, community, and public health settings, including epidemic response and control, little is known about the use of mobile apps in the mental health components of natural disaster management. This scoping review aims to fill this gap in the literature by mapping where and how mobile apps have been used as part of natural disaster mental health response strategies.

This scoping review was reported according to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) [ 28 ]. The PRISMA-ScR checklist is available in Multimedia Appendix 1 . The protocol was not registered.

Search Strategy

A medical librarian (JYK) collaborated with the research team to identify relevant keywords and subject headings for the review, such as mHealth or m-health; mobile health or mobile applications; public health emergency, disaster, or catastrophe; and flood, earthquake, or hurricane. Equipped with this knowledge, the librarian developed and executed comprehensive searches in 6 electronic databases, including Ovid MEDLINE, Ovid Embase, APA PsycInfo, CINAHL, Scopus, and Web of Science Core Collection. The search was conducted on June 30, 2022, and was limited to the English language. The full search strategies are available in Multimedia Appendix 2 .

Inclusion and Exclusion Criteria

We included papers that applied mobile apps for mental health responses to disasters. Papers were included if the study participants were persons affected by a natural disaster (setting), the intervention included using a mobile phone app, and the outcome included the assessment of a mental health problem. Studies in which participants were exposed to a man-made disaster were included if the sample also included some participants exposed to a natural disaster. The mental health conditions included were stress, anxiety, depression, and PTSD. Only full-text studies published in English were included. Studies that did not include any intervention with a mobile app for mental health, those focused on videoconferencing or phone calls, and papers on protocols, trial registration, or review were excluded.

Selection of Studies

The search identified papers that were retrieved from the databases. After removing duplicates, the initial titles and abstracts of the unique papers were screened by 2 independent review authors based on the inclusion criteria in a web-based tool called Covidence (Veritas Health Innovation Ltd) [ 29 ]. Full texts of the selected papers that met the inclusion criteria were reviewed by the 2 independent reviewers. The research team resolved disagreements through discussion. The bibliographies from the included studies were also reviewed to identify additional studies for inclusion.

Data Extraction and Synthesis

Data from each selected full-text paper were extracted into a data extraction form developed by the research team. The data included the author and year of publication, country of study, study design, number of participants, type of natural disaster, name of the mobile app, duration of use of the app, outcome measures, and the study’s findings. These data were synthesized using a narrative approach based on the outcome measures, the duration, frequency of use of the mobile apps, and the outcomes.

Search Results

Of the 1532 papers retrieved from the searches, 976 unique papers had their titles and abstracts screened after deduplication. A total of 38 papers were moved to full-text screening, and data were extracted from 5 papers [ 30 - 34 ] ( Figure 1 ). Table 1 shows the summary of the details of the papers.

a TLS: Training for Life Skills.

b PTSD: posttraumatic stress disorder.

c MBSR: Mindfulness-Based Stress Reduction.

d PFA: Psychological First Aid.

e SAMHSA: Substance Abuse and Mental Health Services Administration.

Characteristics of Included Studies

Of the 5 studies included in this review, 3 (60%) were conducted in the United States [ 30 , 31 , 34 ], while 2 (40%) were conducted in South Korea [ 32 , 33 ]. All studies used different study designs. A total of 3 studies used a quasi-experimental design—the first, a single group postexperiment with 22 participants [ 32 ]; the second, a multiple-baseline single case experimental design with 7 participants [ 30 ], while the third study used a 1-group pre- and posttest design with 318 participants [ 31 ]. The Training for Life Skills (TLS) app study had only a posttest following the use of the app [ 32 ]; the other 2 had baseline and follow-up measurements with the Sonoma Rises app study having, in addition, preintervention and postintervention measurements. The Psychological First Aid (PFA) study was designed as a qualitative study, while the Substance Abuse and Mental Health Services Administration (SAMHSA) study used a mixed methods descriptive design.

Characteristics of the Population

The TLS, Sonoma, and Headspace apps were designed for disaster survivors, while the PFA and SAMHA apps were designed to support disaster relief workers. The TLS app study was administered to adults with a median age of 32 years. Participants of the Sonoma Rises app study had a mean age of 16 (SD 0.98) years, while participants of the Headspace app study had a mean age of 46.1 (SD 10) years. The TLS app study focused on all types of disasters; the Sonoma Rises study focused on adolescents exposed to wildfires, while the Headspace app focused on women who experienced hurricanes and deep-water oil spillage. The PFA study involved 19 disaster health care workers who first underwent disaster simulation training using the mobile app.

Characteristics of the Mobile App Interventions

The included studies revealed several mobile phone apps used as interventions. The first, the TLS app, was used as a psychological first aid program for disaster survivors with content on information, psychological healing, and mood change [ 32 ]. The second was the Sonoma Rises app, a Health Insurance Portability and Accountability Act (HIPAA)–compliant, cloud-based mobile app with daily push notifications as reminders designed to help survivors of wildfires or other disasters to find their new routines, build resilience, and increase well-being. The app included 6 self-paced content sections, psychoeducation, and direct connections to free and local mental health care services. The third was the Headspace app for a mindfulness-based stress reduction program that included a series consisting of 10 sessions designed to be used for about 10 minutes per day. The SAMHSA Disaster App equips behavioral health providers to respond to all kinds of traumatic incidents by enabling them to readily access disaster-specific information and other important materials directly on their mobile devices [ 34 ]. The PFA mobile app provided evidence-based information and tools for disaster workers to prepare for, execute, and recover from providing psychological first aid during disasters. Accessibility via smartphones and the inclusion of multimedia interventions and assessments tailored for disaster contexts were key features enabling its use integrated with the simulation training [ 33 ].

Frequency and Duration of App Use

The 3 survivor-based apps had variations in the duration of the intervention (app use), which were 8 weeks, at least 5 times a week, frequency of use per day not specified [ 32 ]; 4 weeks for 10 minutes per day [ 30 ]; and 6 weeks for 5-10 minutes per day [ 31 ]. Both the TLS app and the Sonoma Rises app studies had weekly follow-up assessments. The different interventions were applied at least a year following the disasters. Participants in the Sonoma Rises app study used the app on an average of 17 (SD 8.92) days and visited the app an average of 43.50 (SD 30.56) times, with an average session lasting 56.85 (SD 27.87) seconds. The mean time spent on the app was 35.77 (SD 30.03) minutes, while for the TLS app study, the median time spent on the app over the 8 weeks of use was 200-399 minutes. Participants used the Headspace app an average of 24 (SD 36) days and logged in an average of 36 (SD 80) times. There was no description of the frequency and duration of use for the relief worker apps.

Effectiveness Outcomes

Effectiveness outcomes refer to the effects or impact of an intervention or program on the intended outcomes or goals. Different measures were used to assess the effectiveness of the apps’ use as either the primary or secondary outcome. Emotional quotients (emotional stability), basic rhythm quotients (brain stability), alpha-blocking rates (increased positive mood), and brain quotients assessed using electroencephalogram (EEG)–measured brainwave activities adjusted for self-reported app use time were used in the TLS app study [ 32 ]. The Headspace app study assessed effectiveness using a combination of measures such as trait mindfulness using a 15-item Mindful Attention Awareness Scale (MAAS)—trait version; depressive symptoms using the Center for Epidemiologic Studies Depression Scale-10 (CESD-10); perceived stress with the Perceived Stress Scale, 4-item version (PSS-4); and sleep quality using the Pittsburgh Sleep Quality Index (PSQI) [ 31 ]. The Sonoma Rises app study measured efficacy using daily ratings of anxiety and fear, weekly measures of post-traumatic stress symptoms using the Child PTSD Symptom Scale (CPSS-5) for Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition ( DSM-5 ), internalizing and externalizing symptoms using the Behaviour and Feelings Survey (BFS), psychosocial functioning using the Ohio Scale for Youth—Functioning subscale (OSY), and measures of anxiety (Generalized Anxiety Disorder-7 [GAD-7]), depression (Patient Health Questionnaire-9 [PHQ-9]), well-being—Warwick-Edinburgh Mental Well-being Scale (WEMWBS), sleep (Insomnia—Severity Index [ISI]), academic engagement (Student Engagement Instrument [SEI]), and perceived social support (Wills’ Social Support Scale [WSSS]) [ 30 ].

All 3 survivor-based apps were found to have positive benefits in addressing mental health issues among persons exposed to natural disasters. The TLS mobile app was shown to be effective in increasing positive and decreasing negative psychological factors according to app use time. The TLS mobile apps’ use had a significant effect on the emotional quotients (β=.550; P <.008), explanatory power (EP) was 30%, had a significant positive effect on the basic rhythm quotient (left brain: β=.598; P <.003; EP 35; right brain: β=.451; P <.035; EP 20%). Additionally, it had a significant positive effect on the alpha-blocking rate (left brain: β=.510; P <.015; EP 26%; right brain: β=.463; P <.035, EP 21%); and a significant positive effect on the brain quotient (β=.451; P <.035; EP 20%) [ 16 ]. The Headspace app had a positive effect on depression (odds ratio [OR] 0.3, 95% CI 0.11-0.81), physical activity (OR 2.8, 95% CI 1.0-7.8), sleep latency (OR 0.3, 95% CI 0.11-0.81), sleep duration (OR 0.3, 95% CI 0.07-0.86), and sleep quality (OR 0.1, 95% CI 0.02-0.96); however, there was no change in mindfulness scores from baseline to follow-up. For the Sonoma Rises app, no significant effects were observed for the clinical and functional outcomes because the longitudinal part of the study was affected by limited statistical power as a result of small sample size and historical confounds that made the participants miss data submission. However, visual inspection of individual data following the intervention showed downward trends across the study phases for daily levels of anxiety, fearfulness, and individual posttraumatic stress symptom severity.

For the PFA app, the qualitative study explored disaster health workers’ experiences with simulation training using focus group discussions. A total of 19 participants engaged in disaster scenarios with standardized patients, using a PFA app for guidance. Workers valued the practical educational approach, felt increased self-efficacy to support survivors, and identified areas for enhancing simulations and app tools to optimize effectiveness.

Implementation Outcomes

Implementation outcomes refer to the effects of an intervention or program implementation on various aspects of the implementation process, such as the fidelity of implementation, acceptability, adoption, feasibility, and maintainability. In the papers reviewed, feasibility was assessed using enrollment, program participation, and retention. Acceptability was measured using how well participants liked the app using a rating scale, how much of the app program was completed, the biggest barriers, and whether the app would be recommended to others. Data on characteristics of app use (engagement) were measured using the total number of log ins, average log ins per program completer, platform used (iOS, Android, or web-based), day of week of use (weekday vs weekend), and time of day of use (in 4-hour blocks) [ 30 , 31 ].

The Headspace app was reported to be cost-effective to implement and easy to use [ 31 ]. For engagement, only 14% (43/318) of the enrolled women used the app. The level of engagement with the app was high, with 72% (31/43) of participants completing some or all the sessions. Retention was also high with 74% (32/43) of the participants completing the follow-up survey. Lack of time was cited as the main barrier to using the app for 37% (16/43) of users and 49% (94/193) of nonusers. The majority of the users (32/43, 74%) reported high levels of satisfaction with the app. Acceptability was also high, with most participants (32/43, 74%) reporting that they liked the app and 86% (37/43) reporting that they would recommend it to others. Characteristics of app use showed that of the 1530 log ins, most participants (n=1191, 78%) used the iOS platform, mainly on weekdays (n=1147, 75%) and at different times of day mostly from noon to 4 PM (n=375, 25%).

Sonoma Rises was found to be feasible in terms of engagement and satisfaction among teens with high levels of disaster-related posttraumatic stress symptoms [ 30 ]. The self-assessment and data visualization features of the Sonoma Rises app strongly appealed to all the participants, and they were willing to recommend the app to their friends. Self-satisfaction with the mobile app was rated as extremely high (mean 8.50, SD 0.58, on a scale of 0 to 10, with 10 as totally satisfied). The participants agreed or strongly agreed to recommend this intervention to a friend. The participants found the intervention helpful (mean 2, SD 0.82); had the content, functions, and capabilities they needed (mean 3, SD 1.12); and were satisfied with how easy it was to use the app (mean 2, SD 0), on a scale of 1 to 5 with 1 as strongly agree and 5 as strongly disagree. In the qualitative feedback, to make the use of the app better, the participants suggested more notifications to return to the app and the use of the app immediately after a disaster. Implementation outcome was not an objective of the TLS app, hence, none was reported.

Other Mobile Apps With Potential Use in Disasters

Some mobile apps not meeting the inclusion criteria showed promise for supporting mental health in disasters. PTSD Coach provides tools for managing PTSD symptoms [ 35 ]. Though not disaster-specific, its psychoeducation, symptom tracking, and coping strategies could aid survivors. Similarly, COVID Coach was designed to help manage pandemic-related stress and anxiety [ 36 ]. These apps are summarized in Table 2 .

a PTSD: posttraumatic stress disorder.

Principal Findings

This review sought to identify and map the use of mobile apps for the mental health component of natural disaster management. We found only 5 studies meeting the inclusion criteria. The scarcity of published literature in this area suggests that mobile apps have not been extensively used in mental health responses to natural disasters. Academic studies on the public’s use of mobile technologies in disaster management are still nascent [ 37 ], but there has been increased interest in developing and deploying digital technology and mobile apps by governments and nonstate actors as part of disaster preparedness and response [ 38 , 39 ]. A recent systematic review found that there is a lack of mental health preparedness in most countries when it comes to disasters [ 40 ]. The 5 studies included in our scoping review confirmed this gap and further demonstrated that mobile apps can provide mental health support to disaster-affected individuals and communities. The studies found that the use of mobile apps was associated with improvements in mental health outcomes, such as decreased anxiety and depression symptoms and increased resilience. The reviewed studies also suggest that mobile apps can be effective in delivering psychoeducation and coping skills training to disaster-affected individuals. A 2017 scoping review found that mobile apps have been largely used for communication purposes in disaster management [ 37 ]. The scope of use was classified into 5 categories which are not mutually exclusive. These categories are (1) crowdsourcing (organize and collect disaster-related data from the crowd), (2) collaborating platforms (serve as a platform for collaboration during disasters), (3) alerting and information (disseminate authorized information before and during disasters), (4) collating (gather, filter, and analyze data to build situation awareness), and (5) notifying (for users to notify others during disasters) [ 37 ].

Some authors classify disaster response into 3 phases: preparedness, response, and mitigation [ 41 ]. The studies included in this review exclusively examined the use of mobile apps during the recovery phase of disaster management. However, none of the studies explored the potential of mobile apps during the preparedness or response phases of disaster management. By addressing this gap, future research could help to provide more comprehensive and effective strategies for the use of mobile apps throughout all phases of disaster management. Examples of potential opportunities are demonstrated in Figure 2 .

Preparedness Phase

Mobile apps can play a critical role as primary prevention interventions by raising awareness and promoting mental health literacy in the community in preparation for natural disasters. These apps can provide information on common mental health problems that may arise during and after disasters and offer tips on staying mentally healthy. For example, apps can include psychoeducation modules on coping skills, stress reduction, and self-care techniques, as well as information on how to prepare for a disaster and what steps to take to protect one’s mental health during and after a disaster. The use and effectiveness of mobile apps in health literacy have been demonstrated in the literature [ 19 ], thus providing a foundation for adaptation in disaster management.

Response Phase

Mobile apps can be used to connect people in need of mental health support with mental health professionals or other resources. For example, apps can provide information on emergency hotlines, crisis intervention services, and support groups. This was demonstrated as effective during the COVID-19 pandemic [ 42 ]. Mobile apps can also provide coping strategies and techniques to manage stress and anxiety in response to other natural disasters [ 34 ]. In this scoping review, we found that 3 apps had positive benefits in addressing mental health issues among persons exposed to natural disasters.

Recovery Phase

As part of secondary and tertiary prevention strategies, mobile apps can provide valuable ongoing support to those affected by disasters. For secondary prevention, mobile apps can be designed to support early detection and intervention for mental health problems after a natural disaster. These apps can include screening tools to identify common mental health issues such as anxiety, depression, and PTSD and offer appropriate referral pathways [ 43 ]. Additionally, apps can provide symptom-tracking tools to help individuals monitor their mental health over time [ 43 ]. For tertiary prevention, mobile apps can support the ongoing management of established mental health problems after a natural disaster. For example, apps can provide evidence-based psychotherapy interventions, such as cognitive-behavioral therapy, to help individuals manage their symptoms [ 44 ]. They can also connect individuals with support groups and peer-to-peer networks to provide additional emotional support and help individuals connect with others who have experienced similar challenges. Furthermore, mobile apps can offer self-help tools, such as meditation exercises and mood tracking, to help people cope with the ongoing mental health effects of the disaster. They can also provide information on local mental health services and support groups, helping individuals access the resources they need to manage their mental health.

General Mental Health Apps Show Promise for Disaster Response

While not specifically designed for disaster contexts, some mobile apps demonstrate strategies to support mental health that could aid disaster survivors. PTSD Coach delivers PTSD psychoeducation, symptom tracking tools, coping skills training, and crisis resource access—elements that could help survivors experiencing common postdisaster issues like trauma or loss [ 35 ]. Though it was tailored for veterans and civilians with PTSD, 1 study found it improved users’ depression and functioning. Similarly, COVID Coach offered pandemic-related stress management through symptom tracking, healthy coping recommendations, and crisis line referrals [ 36 ]. By leveraging the scalability of mobile apps, COVID Coach reached many struggling during a global crisis. These examples illustrate that apps may provide accessible, far-reaching mediums for disseminating disaster mental health resources—even without disaster-specific tailoring. Research should further explore adapting evidence-based, general mental health apps for disaster contexts or incorporate elements of them into future disaster response tools. With mental health needs magnified during disasters, mobile apps with thoughtful design show promise in expanding access to psychosocial support.

There are several potential limitations when using mobile apps for mental health responses to disasters. One of the main concerns is the accessibility of these apps, as not all members of the affected communities may have access to smartphones or internet connectivity. Furthermore, language and cultural barriers may prevent effective use. Another potential limitation is the quality and accuracy of the information provided. Without proper oversight, some apps may provide misinformation or inaccurate advice, which could exacerbate mental health issues. In addition, privacy concerns around collecting and storing sensitive data must be addressed.

Barriers like lack of mobile devices and internet access can impede adoption, especially in marginalized areas. Apps not designed for low literacy users or that are only available in certain languages could also limit accessibility. Concerns around privacy and security may deter some individuals. However, smartphone ubiquity globally enables use by vulnerable groups. Government agencies and nongovernmental organizations (NGOs) can promote adoption by integrating vetted apps into disaster protocols and funding dissemination. Developing apps with stakeholders and prelaunch user testing also facilitate uptake. Monitoring user feedback allows for ongoing optimization and troubleshooting of barriers. Cultural tailoring to address stigma and use local beliefs further enables implementation success. Finally, limited evidence-based research into app effectiveness highlights the need for more rigorous evaluation and testing of mobile apps for disaster mental health response.

This scoping review has certain methodological limitations that should be considered while interpreting its results. First, the search was restricted to 6 electronic databases and only English-language papers were considered. We also searched MEDLINE and not PubMed, and these may have led to the omission of some relevant studies. Second, the study focused on mobile phone apps for mental health response to disasters, disregarding other types of technology that could also be used in disaster management such as telehealth, SMS text messaging, and emails. Moreover, since the study included only 5 papers, it may not offer a comprehensive overview of the use of mobile phone apps in disaster response strategies. There is the possibility of the existence of apps not yet published in academic literature. Fourth, the nonuse of a control group in the design of the studies makes it difficult to determine whether the observed effects were entirely due to the use of the apps or other characteristics of the participants that predisposed them to use the apps. Fifth, the small sample sizes for the studies mean they require caution with generalization. Despite these limitations, the review provides valuable insights into the use of mobile apps in disaster response and serves as a useful resource for developing contextually appropriate mobile apps for disaster management. Last, our study focused on natural disasters, further research should examine the role of apps in supporting mental health in conflict and complex emergencies such as wars, outbreaks of violence, and complex political conflict situations [ 45 ].

Conclusions

This scoping review found that mobile apps have not been extensively used in mental health responses to natural disasters, with only 5 studies meeting the inclusion criteria. However, the studies included in this review demonstrate that mobile apps can be useful in providing mental health support to disaster-affected individuals, as well as equip disaster responders. There is a critical gap identified in this study, as none of the studies investigated the use of mobile apps for potential victims in the preparedness or response phases of disaster management. We, therefore, recommend that mobile apps be integrated into the various phases of disaster management as part of mental health response. Additionally, it is important to ensure that these apps are accessible to all members of the community, taking into account cultural, linguistic, and other factors that may impact their effectiveness. Mobile apps have great potential to provide valuable ongoing support to those affected by disasters, and they can be a valuable resource in disaster management, helping people cope with the mental health effects of disasters and connecting with the necessary support services.

The findings from this scoping review have important implications for policy makers, disaster management professionals, and mental health practitioners. There is a clear need for policies and protocols that integrate evidence-based mobile apps into mental health disaster planning and response. Disaster agencies should invest in developing, evaluating, and widely disseminating mobile apps specifically designed to mitigate psychological trauma before, during, and after catastrophic events. Mental health professionals can incorporate vetted mobile apps into their standard of care for at-risk disaster survivors. Going forward, a collaborative approach across these groups will be essential to leverage mobile technology in building community resilience and addressing the rising mental health burdens in an era defined by climate change–fueled natural disasters.

Acknowledgments

This work was funded by the Department of Psychiatry, Dalhousie University, Halifax, Canada. The funder was not involved in the conceptualization or implementation of the study, nor the decision to publish the findings.

Conflicts of Interest

None declared.

The PRISMA-SCR checklist. PRISMA-SCR: Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews.

Detailed search strategy.

• Birkmann J, Cardona OD, Carreño ML, Barbat AH, Pelling M, Schneiderbauer S, et al. Theoretical and conceptual framework for the assessment of vulnerability to natural hazards and climate change in Europe. In: Birkmann J, Kienberger S, Alexander DE, editors. Assessment of Vulnerability to Natural Hazards: A European Perspective. San Diego, California. Elsevier; 2014;1-19.
• Bhola V, Hertelendy A, Hart A, Adnan SB, Ciottone G. Escalating costs of billion-dollar disasters in the US: climate change necessitates disaster risk reduction. J Clim Change Health. 2023;10:100201. [ CrossRef ]
• Leppold C, Gibbs L, Block K, Reifels L, Quinn P. Public health implications of multiple disaster exposures. Lancet Public Health. 2022;7(3):e274-e286. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Global SDG indicators data platform. United Nations. 2021. URL: https://unstats.un.org/sdgs/dataportal [accessed 2024-04-05]
• Sloggy MR, Suter JF, Rad MR, Manning DT, Goemans C. Changing climate, changing minds? The effects of natural disasters on public perceptions of climate change. Clim Change. 2021;168(3-4):25. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Chen X, Bakal J, Whitten T, Waldie B, Ho C, Wright P, et al. Assessing the impact of COVID-19 pandemic on the health of residents and the healthcare system in Alberta, Canada: an observational study-the Alberta post-COVID follow-up study. BMJ Open. 2023;13(2):e067449. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Pazderka H, Shalaby R, Eboreime E, Mao W, Obuobi-Donkor G, Agyapong B, et al. Isolation, economic precarity, and previous mental health issues as predictors of PTSD status in females living in Fort McMurray during COVID-19. Front Psychiatry. 2022;13:837713. [ FREE Full text ] [ CrossRef ] [ Medline ]
• To P, Eboreime E, Agyapong VIO. The impact of wildfires on mental health: a scoping review. Behav Sci (Basel). 2021;11(9):126. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Agyapong B, Shalaby R, Eboreime E, Obuobi-Donkor G, Owusu E, Adu MK, et al. Cumulative trauma from multiple natural disasters increases mental health burden on residents of Fort McMurray. Eur J Psychotraumatol. 2022;13(1):2059999. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Guerra O, Eboreime E. The impact of economic recessions on depression, anxiety, and trauma-related disorders and illness outcomes-a scoping review. Behav Sci (Basel). 2021;11(9):119. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Norris FH, Friedman MJ, Watson PJ, Byrne CM, Diaz E, Kaniasty K. 60,000 disaster victims speak: part I. An empirical review of the empirical literature, 1981-2001. Psychiatry. 2002;65(3):207-239. [ CrossRef ] [ Medline ]
• North CS, Pfefferbaum B. Mental health response to community disasters: a systematic review. JAMA. 2013;310(5):507-518. [ CrossRef ] [ Medline ]
• Pfefferbaum B, Flynn BW, Schonfeld D, Brown LM, Jacobs GA, Dodgen D, et al. The integration of mental and behavioral health into disaster preparedness, response, and recovery. Disaster Med Public Health Prep. 2012;6(1):60-66. [ CrossRef ] [ Medline ]
• Math SB, Nirmala MC, Moirangthem S, Kumar NC. Disaster management: mental health perspective. Indian J Psychol Med. 2015;37(3):261-271. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Ali A, Katz DL. Disease prevention and health promotion: how integrative medicine fits. Am J Prev Med. 2015;49(5 Suppl 3):S230-s240. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Reddy MC, Paul SA, Abraham J, McNeese M, DeFlitch C, Yen J. Challenges to effective crisis management: using information and communication technologies to coordinate emergency medical services and emergency department teams. Int J Med Inform. 2009;78(4):259-269. [ CrossRef ] [ Medline ]
• Cheng DR, Coote A, South M. A digital approach in the rapid response to COVID-19—experience of a paediatric institution. Int J Med Inform. 2021;149:104407. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Souza F, Kushchu I. Mobile disaster management system applications—current overview and future potential. In: IOP Conference Series: Materials Science and Engineering. 2005. Presented at: Proceedings EURO mGOV; 2005;455-466; NA. URL: https://iopscience.iop.org/article/10.1088/1757-899X/1009/1/012049
• Chandrashekar P. Do mental health mobile apps work: evidence and recommendations for designing high-efficacy mental health mobile apps. Mhealth. 2018;4:6. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Hyde AM, Watt M, Carbonneau M, Eboreime EA, Abraldes JG, Tandon P. Understanding preferences toward virtual care: a pre-COVID mixed methods study exploring the perspectives of patients with chronic liver disease. Telemed J E Health. 2022;28(3):407-414. [ CrossRef ] [ Medline ]
• Eboreime E, Ohinmaa A, Rusak B, Cassidy KL, Morrison J, McGrath P, et al. The Text4HealthyAging program: an evidence-based text messaging innovation to support healthy urban aging in Canada and Australia. Gerontol Geriatr Med. 2022;8:23337214221081378. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Shalaby R, Adu MK, Andreychuk T, Eboreime E, Gusnowski A, Vuong W, et al. Prevalence, demographic, and clinical correlates of likely PTSD in subscribers of Text4Hope during the COVID-19 pandemic. Int J Environ Res Public Health. 2021;18(12):6227. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Vilendrer S, Amano A, Johnson CGB, Favet M, Safaeinili N, Villasenor J, et al. An app-based intervention to support first responders and essential workers during the COVID-19 pandemic: needs assessment and mixed methods implementation study. J Med Internet Res. 2021;23(5):e26573. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Obuobi-Donkor G, Eboreime E, Bond J, Phung N, Eyben S, Hayward J, et al. An E-mental health solution to prevent and manage posttraumatic stress injuries among first responders in Alberta: protocol for the implementation and evaluation of text messaging services (Text4PTSI and Text4Wellbeing). JMIR Res Protoc. 2022;11(4):e30680. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Eboreime E, Shalaby R, Mao W, Owusu E, Vuong W, Surood S, et al. Reducing readmission rates for individuals discharged from acute psychiatric care in Alberta using peer and text message support: protocol for an innovative supportive program. BMC Health Serv Res. 2022;22(1):332. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Andersson G, Cuijpers P. Internet-based and other computerized psychological treatments for adult depression: a meta-analysis. Cogn Behav Ther. 2009;38(4):196-205. [ CrossRef ] [ Medline ]
• Nicholas J, Ringland KE, Graham AK, Knapp AA, Lattie EG, Kwasny MJ, et al. Stepping up: predictors of 'stepping' within an iCBT stepped-care intervention for depression. Int J Environ Res Public Health. 2019;16(23):4689. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Tricco AC, Lillie E, Zarin W, O'Brien KK, Colquhoun H, Levac D, et al. PRISMA extension for Scoping Reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med. 2018;169(7):467-473. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Covidence systematic review software. Veritas Health Innovation. Melbourne, Australia.; 2022. URL: https://www.covidence.org/ [accessed 2024-04-05]
• Heinz AJ, Wiltsey-Stirman S, Jaworski BK, Sharin T, Rhodes L, Steinmetz S, et al. Feasibility and preliminary efficacy of a public mobile app to reduce symptoms of postdisaster distress in adolescent wildfire survivors: Sonoma rises. Psychol Serv. 2021;19(2):67-79. [ CrossRef ] [ Medline ]
• Rung AL, Oral E, Berghammer L, Peters ES. Feasibility and acceptability of a mobile mindfulness meditation intervention among women: intervention study. JMIR Mhealth Uhealth. 2020;8(6):e15943. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Choi YJ, Ko EJ, Choi EJ, Um YJ. Managing traumatic stress using a mental health care mobile app: a pilot study. Int J Ment Health Promot. 2021;23(3):385-393. [ FREE Full text ] [ CrossRef ]
• Choi YJ, Jung HS, Choi EJ, Ko E. Disaster healthcare workers' experience of using the psychological first aid mobile app during disaster simulation training. Disaster Med Public Health Prep. 2021;17:e55. [ CrossRef ] [ Medline ]
• Seligman J, Felder SS, Robinson ME. Substance Abuse and Mental Health Services Administration (SAMHSA) behavioral health disaster response app. Disaster Med Public Health Prep. 2015;9(5):516-518. [ CrossRef ] [ Medline ]
• Kuhn E, van der Meer C, Owen JE, Hoffman JE, Cash R, Carrese P, et al. PTSD coach around the world. Mhealth. 2018;4:15. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Jaworski BK, Taylor K, Ramsey KM, Heinz A, Steinmetz S, Pagano I, et al. Exploring usage of COVID coach, a public mental health app designed for the COVID-19 pandemic: evaluation of analytics data. J Med Internet Res. 2021;23(3):e26559. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Tan ML, Prasanna R, Stock K, Hudson-Doyle E, Leonard G, Johnston D. Mobile applications in crisis informatics literature: a systematic review. Int J Disaster Risk Reduct. 2017;24:297-311. [ CrossRef ]
• Romano M, Onorati T, Aedo I, Diaz P. Designing mobile applications for emergency response: citizens acting as human sensors. Sensors (Basel). 2016;16(3):406. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Schoning J, Rohs M, Krüger A, Stasch C. Improving the communication of spatial information in crisis response by combining paper maps and mobile devices. In: Mobile Response 2008. Lecture Notes in Computer Science, vol 5424. Berlin, Heidelberg. Springer; 2009;57-65.
• Roudini J, Khankeh HR, Witruk E. Disaster mental health preparedness in the community: a systematic review study. Health Psychol Open. 2017;4(1):2055102917711307. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Colajanni G, Daniele P, Nagurney A, Nagurney LS, Sciacca D. A three-stage stochastic optimization model integrating 5G technology and UAVs for disaster management. J Glob Optim. 2023;86:1-40. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Singh HJL, Couch D, Yap K. Mobile health apps that help with COVID-19 management: scoping review. JMIR Nurs. 2020;3(1):e20596. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Funnell EL, Spadaro B, Martin-Key N, Metcalfe T, Bahn S. mHealth solutions for mental health screening and diagnosis: a review of app user perspectives using sentiment and thematic analysis. Front Psychiatry. 2022;13:857304. [ FREE Full text ] [ CrossRef ] [ Medline ]
• Murphy K, Coombes I, McMillan S, Wheeler AJ. Clozapine and shared care: the consumer experience. Aust J Prim Health. 2018;24(6):455-462. [ CrossRef ] [ Medline ]
• Goniewicz K, Burkle FM, Hertelendy AJ, Khorram-Manesh A. The impact of war on emergency departments visits by Ukrainian refugees in Poland. Am J Emerg Med. 2023;67:189-190. [ CrossRef ] [ Medline ]

Abbreviations

Edited by G Eysenbach; submitted 13.06.23; peer-reviewed by T Benham, K Goniewicz, R Konu, J Ranse, P Moreno-Peral; comments to author 10.01.24; revised version received 25.02.24; accepted 23.03.24; published 17.04.24.

©Nwamaka Alexandra Ezeonu, Attila J Hertelendy, Medard Kofi Adu, Janice Y Kung, Ijeoma Uchenna Itanyi, Raquel da Luz Dias, Belinda Agyapong, Petra Hertelendy, Francis Ohanyido, Vincent Israel Opoku Agyapong, Ejemai Eboreime. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 17.04.2024.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.