Student-to-student writing guide part 2: the sections of a scientific manuscript

“heavy lifting” by JonathanCohen is marked with CC BY-NC 2.0
“heavy lifting” by JonathanCohen is marked with CC BY-NC 2.0

There are lots of different compelling ways to tell scientific stories. Ultimately they all have to be wrestled into the same standardized framework, the one that we’ve come to expect when we read a paper: Abstract and Introduction, then Results and Discussion, then Methods and References. Different journals and article formats will have different constraints on the formatting of each of these sections, and some may even combine sections or require expanded versions of sections in the supplement. Before you write, you will want to have a look at the Guidelines for Authors page for a few journals that you might submit to. There is good advice on how to approach some of the sections of a typical paper on Tony Hyman’s lab website. I add my own advice on these sections below.


It’s difficult to provide specific advice on writing abstracts because the Abstract takes a different form and serves a different purpose for different journals and article formats. For some, it is expected to be a standalone summary of the manuscript, in which case you’re usually given ample space (200-250 words) to outline the major question, findings, and conclusions of the manuscript. For others, it’s meant to be something more akin to a lead-in paragraph for the introduction; in these cases the word limits are typically more restrictive, closer to 150 words.

At a minimum, the Abstract for a research study should: state the research question, give background essential to understanding the research question, briefly describe the approach taken, concisely report experimental results, and explain how the new results answer the research question and move the field forward. How much detail you provide for each section of the Abstract will depend on what kind of abstract you are writing, how much space you have, and the objective of the manuscript. For example, for a methods paper, the description of your approach may be expanded, and the description of results may be very brief. An example of an abstract with annotations added to make it clear what each sentence accomplishes is provided in the formatting guide for the journal Nature.

It’s key to remember that many people will only read the Abstract of your manuscript. For this reason, you’d like the Abstract to effectively recapitulate the entire manuscript. Don’t spend a lot of words talking about something that didn’t get a lot of words in the main manuscript. Communicate the main findings clearly, using forceful language to highlight what is truly “new” in your study. The best abstracts communicate why a manuscript is worth reading while simultaneously eliminating the need to read it for those who don’t care about the details.


The goal of the Introduction is to give enough background that your reader can understand what motivated your research and why it is interesting, timely, and important. The goal is not to provide a comprehensive review of the literature on your research subject. If you try to do that you will bury readers with what will seem to them like random facts, and you will lose them quickly.

A statement motivating your research can take many forms, but a good one will make the entire manuscript come across more clearly by setting you up to make conclusions. The perfect research question cannot be answered by any single result but can be answered by the sum total of all of your results. If your question is too narrow, your reader will wonder why they should read your whole paper to get the answer, and they’ll decide instead to skim the figures until they’re satisfied that they get the gist. If the question is too broad, you won’t be able to “declare victory.” Your manuscript will read like an incremental advance toward a hopelessly distant goal, a list of results that means next to nothing to anyone not steeped in the literature of your field. A properly framed research question gives you the opportunity to interpret your results and come to an appropriate conclusion. This will make it easier to write (and read!) the Results and Discussion sections, because you and your reader will have an idea of what the work is building toward. A manuscript with conclusions is satisfying, so you should set yourself up to make conclusions.

The rest of the Introduction should build to the presentation of your research motivation. Citing appropriate literature, explain your experimental system and describe what is known about it, leading your reader through previous research until they have an up-to-date working model. If you do this well, your readers will infer your research question, because you have provided the foundation through literature review. The Introduction acquaints the reader with your research question, while your research question motivates your experiments.

Finally, at the end of the Introduction, concisely state specifically what you set out to do, and then summarize your results with one or two sentences. The Introduction is an odd place to put a conclusion, but it serves two purposes here. First, you could lose readers at any point. If you have a conclusion at the end of the Introduction, anyone who quits reading at that point will still know what you think you found. Second, if your reader knows what you are building toward, they will more easily understand how you are interpreting your results, whether or not they agree with your interpretation. This signpost signifies that the review of accepted facts portion of the manuscript is ending and the start of experimental results and interpretations is ahead.


The point of the Results section is to describe the experiments that gave rise to each figure panel. You should explain what each experiment aims to accomplish and how that relates to what you’ve told the reader so far. You should also outline, in broad strokes, how the experiment was carried out. Experimental details belong in the Materials and Methods section, but in the Results section, you should describe what was done and why.

A simple way to structure the Results section is by alternating between paragraphs outlining how and why experiments were done and paragraphs that describe the results of those experiments. Topic sentences for paragraphs outlining how and why experiments were done can resemble the form: “to determine [crucial piece of information X], we did [experiment Y].” The body of the paragraph can explain why you needed [crucial piece of information X], how you conducted [experiment Y], or can do both. Paragraphs describing results should directly follow the experimental explanation paragraphs. Their topic sentences can resemble the form: “When we did [experiment Y] we found [crucial piece of information X].” The body of the paragraph can then explain what you observed when you did the experiment and should refer to all figure panels showing data from the experiment described in the preceding paragraph. From time to time, you may choose to interleave a paragraph transitioning from one set of experiments (e.g., those contained in a single figure) to another set (e.g., contained in the next figure). Keep alternating paragraphs like this until you’ve explained all of your figure panels. By alternating paragraphs in this way, you will clearly explain your figures while keeping individual paragraphs short enough for your reader to easily parse.


The Discussion section can be the most intimidating part of a manuscript to write because its purpose is the least specific and its structure is not as stereotyped as the other sections. The goal in the Discussion is to frame the results of the study in the context of the larger field and to highlight open questions for future studies. That can be a daunting task. It’s tempting to open the Discussion with a mini-review article summarizing the entire history of the field, but this is a mistake. Essential background information belongs in the Introduction, so repeating essential background information in the Discussion is wasting valuable manuscript lines repeating yourself. You want your reader to be thinking about the meaning of the results of your study, so you should keep the spotlight on your results, even in the Discussion.

The truth is that there are many ways to write a compelling Discussion, but one of my PhD committee members gave me advice on how to structure the Discussion that made so much sense to me that I feel it would be criminal not to share it. To keep the focus of your writing on your results, make the topic sentence of each paragraph of the Discussion section a single conclusion from your study. Then, in the body of the paragraph, discuss that conclusion in the context of what was already known. Does the conclusion: build upon or revise our previous understanding? Add a more detailed mechanistic understanding to a process that we already knew about? Reconcile conflicting observations from the past? This is the place to explain, concisely, why the result is important and valuable to the research community. Once you are done, start a new paragraph, with the topic sentence being another conclusion from the study. Write a paragraph for each major conclusion of the study. Once you have done so, your Discussion is done.

Materials and Methods

Writing the Materials and Methods section should be more straightforward than any other section of the manuscript (as long as you have kept detailed notes in your lab notebook). Look back through the Results section and the figures and make a list of all the experiments that you did, then describe the experimental procedure with enough detail that another person in your field could replicate it within a couple of tries. Include detailed information about both materials and equipment as well as step-by-step descriptions of the method itself. The best Materials and Methods sections also make note of what doesn’t work. If an incubation needs to happen for at least some minimum amount of time, but cannot happen for more than some maximum amount of time, it’s courteous and collegial to say so. If your study made use of a complicated technique and a separate methods-focused paper would be useful to a reader trying to replicate the technique, then the better description should be cited. However, these days, the most scholarly journals allow ample or even unlimited space for descriptions of methods, so the best practice is to keep citations to previous Materials and Methods sections to a minimum to save your reader the trouble of falling down a rabbit hole of citations.


Like the Materials and Methods section, the References should be fairly self-explanatory. Once again, most scholarly journals allow unlimited space for references, so you should never exclude an appropriate reference just to shorten your manuscript. On the other hand, there is no need to provide an exhaustive review of the literature in your field, and you should avoid adding sentences or paragraphs to your manuscript solely to cite a paper that “seems relevant.” Just cite the papers that are essential background.

The biggest challenge in the References section can be adhering to formatting guidelines and keeping works cited in the right order. This task is made much easier if you use a citation management software that can automatically generate a bibliography in the style of the journal you plan to submit to. Endnote is licensed by many research institutions. Mendeley is a free PDF reader designed for scientific publications that also has a capable Microsoft Word plugin. Other software options are available, and any citation manager worth its salt should be able to basically complete the References section for you.

In this writing guide, I have done my best to give a boilerplate structure for each section of a typical scientific manuscript that can be filled in with whatever research story you are equipped to tell with your data. There are many ways to write each section of a paper, and the content suggestions and paragraph structures I have outlined here are just one example. But there is no shame in using a well-established framework and filling it in with your own results. After all, at the end of the day, our time is best spent doing experiments, analyzing data, and making discoveries. As long as your writing is clear and organized, it will get your point across! Write your papers in whatever way feels natural, share them with the world, and get back to the bench where the best parts of science happen.

Link to Student-to-student writing guide part 1: the structure and style of a science story

About the Author:

Ross Pedersen (Twitter: @RossTAPedersen) Is a postdoctoral fellow in Yixian Zheng’s lab in the Department of Embryology at the Carnegie Institution for Science in Baltimore, Maryland, where his research aims to elucidate the pathway governing nuclear lamin assembly following mitosis.