Woman Of Science

Writing a Draft Manuscript

I have been working on a manuscript about some pretty nifty science. I have been working on it for some time, and I finally just submitted it last week. As part of the “how to write” series, I thought I would give my personal process for how to write a manuscript. I am sure there are as many ways to write a manuscript as there are manuscripts that have been written and published. This is my way, and it also how I instruct to my students to help them get over the hurtles of writing.

1. Make figures. Science is all about the data. The data is the story, so the first thing I do is make figures. This can be pretty challenging in and of itself. When working with students, I often have them make the figures – or at least the first draft of the figures. They often don’t know exactly what to do – even if they have presented some nice plots and graphs in group meetings. I usually sketch out the figures with them on the white board before I set them to work getting the data into figure form. Also, I try to get the students to think about the figures as they are taking the data. I usually sketch out plots when we discuss the experiments. If you are already thinking about your data in figure format as you take it, making the figures becomes much easier.

If you figures are missing data, if often become clear at this step. It is obvious that something – some data, some plot – is missing in the figures. We don’t go onto the next step until we have all the data in figure form.

Just FYI, I use Illustrator to make my figures. I hate Powerpoint. It does not allow very much control or good resolution. I know some people love Powerpoint for figures, but I think it is clumsy compared to Illustrator.

2. Long figure captions. After the figures are all made – with ALL the data – we write long figure captions. The figure captions include how we did the experiments, what the figure results show, and what the results mean. The point of a long figure caption is to have an easy way to move from the data to the rest of the manuscript. The “how we did it” becomes part of the materials and methods. The “what the figure results show” become the results section. The “what the results mean” becomes part of the discussion section. The point of the long figure captions is just to help students get over their fear of writing. It is much easier to write while you have a figure to look at and to write about.

3. Methods. I find that the methods are often one of the easier parts to write – especially for students. This is where they get to say what they did. One issue I find, especially when working with undergraduate authors, is that this section can be harder to write than it seems. Yes, it is just what you did, but it has a specific style. For instance, it is not helpful to describe the volumes used in your assays – it is important to describe the concentrations of the reagents. The experiment *should* work the same if you mix up 1 ml or 100 ml, as long as the concentrations are the same.  I have a few favorite papers of my own or from others that I think have particularly good methods sections. I often give these to students to read to help them get the style and tone in their heads before they write.

4. Results. I, personally, find the results easiest to write – especially if you have the figures made up already. Often, each figure is a section of the paper’s results section. Again, this is where one can just describe what the data says. I write as stream of conciousness from the images and describe the data. Doing it this way tends to make the paper a bit long-winded, and it needs a lot of editing, but it is better than to stare at a blank page. It is always easier to cut and edit than to come up with perfect words the first time. A lot of times, I end up writing implications or how the data relates to other work in this section. These ideas are really better suited for the discussion, but it is easier to write them here and move them, if they belong somewhere else.

5. Discussion. After writing the results, I write a discussion section. Sometimes, the results and discussion sections are melded together, and for each figure, I write results and then discussion/implications. If they are separate, I still often write discussion-like ideas in the results and just cut and paste them into the discussion afterwards. These are the starting point, and I expand the discussion from there. At this point, I often have to do additional reading of the literature in order to put my work into context. I add in the references as I think of them with some sort of demarkation that my citation software can recognize. I use {curly brackets} and author name and year. If I know I need a citation as I am writing, I might highlight {cite}, so I can go back and find it later.

6. Introduction. I write the introduction after the discussion. How else will I know what topics I need to introduce and background literature to ground the work until I know the results and implications. I often need to do more reading at this point to make sure I have good and correct references. I insert the citations as I describe above in the text as they come up in the introduction. I often have to read other introductions, especially if I am stuck for words. By this point, I have a good idea of where I am sending the manuscript, so I read introductions from published papers in the same journal to get the style in my head.

7. Abstract. After the paper is pretty much written, I then write the abstract. The abstract is difficult because you need to be brief, for some journals less than 150 words! But, you have to get all the information of why it is important, what you did, and why is matters to the field. Again, I often write something much longer and have to cut cut cut. I think of an abstract as an inverted pyramid – start broad and focus down. You also want to hook the reader – tell them why this work is important early on so they want to read the paper.

8. Other stuff. This stuff I kind of do when the whim hits me. If I am having a hard time writing, then I might write the Acknowledgements section. This takes looking up all the funding agencies and getting the numbers. I had it on a sticky note (an electronic one on computer – not an actual sticky note), but I started being more organized with it all in an excel spreadsheet. Something like the title might change as the paper is being written and different important parts come into view. I like it to represent the results or the implications – it depends on the journal. The authors and author order are usually obvious and depend on the amount of effort, type of person, and the field you are in.

9. Cover letter. The cover letter. I think this is very important, and I think I have been doing it wrong until recently. Once you know the story, the cover letter is the place where you sell it to the editor. The higher “impact” the journal, the more important the cover letter is. This is especially true if the scientific editors are professional science editors and not Principle Investigators acting as editors. You have to educate the editor about how important your field is and why your work is an important piece of the puzzle that was missing until now. You have to discuss what your results are in laymen’s terms – like when you write a proposal for generally educated scientists who might not be right in your field. It must be clear and convincing.  I hope my recent cover letter works. It is far superior to any cover letter I have even submitted before now.

10. References. This is the absolutely last thing I do. Many of the references are being inserted in {curly brackets} along the way, but some will still be missing. They will also be missing from my reference software, so I have to spend some time getting all the citations into that, too. I personally do not use EndNote, but many people do. I use a program called Sente. Either way, I have to get the references into this software system to get inserted at the end. Once my references are inserted in the journal’s style, I cannot modify them or add more, so this is why I wait until the very very last to insert them – otherwise, I am wasting my time.

Before putting in final references, I have my entire lab read and edit the paper. They edit for typos, grammar, spelling, and tell me if things are not clear or confusing.

After all these steps, then I spend a lot of time submitting. The submission processes online are actually easier now than when I was a graduate student, but they still take several hours of inputting names by hand and getting the figures, tables, and writing all uploaded. So, even when you are done with the manuscript, you aren’t quite done.

And there it is, now you have a submitted manuscript! Easy, right? No, but it works, and it is relatively painless – at least for me. So, what are your tricks for writing manuscripts? Comment or post here.