Contributed by Michael Hsieh
I’ve had the privilege of writing a few manuscripts in my research career to date, and helping trainees write them. It’s hard work, but planning and organization helps. Here’s some thoughts on how to approach writing manuscripts based on original biomedical research.
Getting ready to write
Involve your principal investigator (PI) early and throughout the process. It’s our job to help you write!
Write down your hypothesis/research question. Everything else will be spun around this.
Gather your proposed figures and tables in a sequence that tells a story. This will form the basis of your Results section. Write bulleted captions for the figures/tables, including a title that explains the key finding for each figure/table, an explanation of experimental groups and associated symbols/labels, and details on biological and technical replicates and statements (such as “one of four representative experiments are shown.”)
Generate a bulleted outline of the major points for each section of the manuscript. This depends on the journal, but typically, and with minor variations: Introduction, Methods, Results, Discussion. Use Endnote, Reference Manager, Mendeley, or other citation software to start inserting references to go with bullets. Decide from the beginning what word processing software you’ll use (Word, Google Docs, etc.). Google Docs can be helpful for maintaining a single version of the manuscript, but citation software often doesn’t play well with Google Docs (whereas most software options can automatically update citation changes in Word). Here’s what should go in each of these sections:
Introduction: What did you study, and why is it important? What is your hypothesis/research question?
Methods: What techniques did you use? Each technique should be its own bullet, with sub-bullets for key details. If you used animal or human subjects, include a bullet on ethics approval. Important methodologies and materials, i.e., blinding for subjective analyses, full names of cell lines/strains/reagents and your commercial/academic sources for them.
Results: What were your findings? Each major finding should be its own bullet, with sub-bullets going into more detail for each major finding. These bullets should refer to your figures.
Discussion: Summarize your findings in the context of prior work. Discuss possible interpretations. It is important to include a bullet describing the limitations of the presented work. Mention possible future directions.
Now read the entire outline (including the figures). Is it a complete story? If so, you’re ready to prepare for submission. If not, you should have a good idea of what it will take to finish the manuscript.
Writing your manuscript
You first need to decide where you want to submit your manuscript. I like to consider my ideal target audience. I also like to vary which journals I publish in, both to broaden the potential readers of my papers and to avoid the appearance of having an unfair “inside connection” to a given journal. Your academic reputation is priceless.
Once you’ve chosen your journal, look at the journal’s article types. Decide which article type you would like to submit and reformat your outline according to the journal’s standards (including citation style).
Convert your outline (including the figure captions) to complete sentences. Don’t focus on writing perfect prose for the first draft. Write your abstract after the first draft is completed. Make sure the manuscript conforms to the target journal’s word and figure limits.
Discuss all possible authors with your PI. If the study involved many people, create a table of possible authors showing their specific contributions to the manuscript. (This is helpful to do in any case as many journals now require this information.) Assigning authorship is sometimes complicated, but keep in mind that the Acknowledgements can be used to recognize those who made minor contributions (including reading the manuscript to provide feedback). “Equal contribution” authorship positions for the first and last authors is a newer option for a number of journals. An alternative is to generate the initial outline or first draft with the help of co-authors. This can take a lot more work and coordination, but may make sense for highly collaborative and large manuscripts.
Decide with your PI who will be corresponding author. Usually you or the PI.
Circulate the manuscript draft to all possible authors. Thank them for their prior and ongoing support. Inform your co-authors where you would like to send the manuscript and why. Give them a reasonable deadline to provide feedback (minimum of a few weeks). If you use Microsoft Word, ask your co-authors to use track changes.
Collate comments from your co-authors. The Combine Documents function in Word can be very helpful. Consider reconciling all comments and tracked changes before circulating another manuscript draft so that co-authors can read a “clean” copy. Repeat this process until you and your PI (and co-authors) are satisfied that the manuscript is ready for submission.
Some prefer to avoid listing authors on manuscript drafts until the final version is generated because the relative contributions of authors can shift during manuscript preparation.
Submit your manuscript
Write a cover letter for your manuscript. Put it on institutional letterhead, if you are permitted by the journal’s submission system. This makes the cover letter, and by extension, the manuscript, more professional. Some journals have required language for cover letters regarding simultaneous submissions to other journals. It’s common for journals to require that cover letters include a rationale explaining the impact and findings of the manuscript. If you need to do this, include key references and a citation list at the end of the cover letter.
Most journals will require you to provide keywords, and/or to choose subject areas related to the manuscript. Be prepared to do so.
Conflicts of interest should be declared in the manuscript, even if the journal does not explicitly request this. Ask your co-authors about any such potential conflicts.
Gather names and official designations of any grants that supported the work described in your manuscript. Ask your co-authors and your PI. This is very important for funding agencies such as the NIH, which scrutinize the productivity of their funded investigators and take this into account when reviewing future grants.
It’s common for journals to allow you to suggest an editor to handle your manuscript. Editors with expertise in your area are more likely to be able to identify and recruit reviewers who are also well-versed in the subject matter of your manuscript. Discuss this with your PI and co-authors.
Likewise, journals often allow authors to suggest reviewers. Some meta-literature indicates that manuscripts with suggested reviewers have an overall higher acceptance rate. It also behooves you to have expert reviewers that can evaluate your manuscript fairly, but also provide feedback that can improve your paper if revisions are recommended. Avoid suggesting reviewers at your own institution or who have recently written papers or been awarded grants with you. Savvy editors look for these types of relationships between reviewers and authors, and will nix a suggested reviewer with any potential conflict of interest. Discuss suggested reviewers with your PI and co-authors.
On the flip side, many journals will allow you to list opposed reviewers. If you believe that someone specific will provide a negatively biased review for non-scientific reasons, that is grounds for opposing them as your manuscript’s reviewer. In small fields, it may not be possible to exclude reviewers and still undergo expert peer review. Definitely a must-discuss with your PI and co-authors.
Generate a final version of the manuscript. Most journals use online submission systems that mandate uploading individual files for the manuscript, cover letter, etc. You may have to use pdf converting software (i.e., Adobe Acrobat) to change Word documents to pdf’s, or to combine documents into a single pdf. Review the final version, including the resolution and appearance of figures. Make sure that no edges of text or graphics near page margins are cut off (Adobe Acrobat sometimes does this with Microsoft Word). Send the final version to your PI and co-authors. Revise any errors. Then submit! Good luck!
Michael Hsieh is the Stirewalt Scientific Director of the Biomedical Research Institute and an Associate Professor at the George Washington University, where he studies host-pathogen interactions in the urinary tract. Michael has published over 90 peer-reviewed scientific papers. His work has been featured on PBS and in the New York Times. ORCID.