Back to Writing for 'Realz'...

One of the projects that I started when I got to Stanford in October last year is finally at a stage where it's time to begin crafting it into a 'story': I've got some interesting results for which I've received a lot of feedback, and at this point figuring out what needs to be done next will best be served by trying to 'put it all together'. Therefore, I've begun writing a manuscript. Note that this doesn't mean that I'll be submitting it anytime soon - I may decide that additional, significant things need to be done - but for now I'm trying to put together a coherent draft.

I began by setting aside an entire day this past weekend to research and write the introduction. As I've mentioned before, the first advice given to me on how to write a paper (from my M.Sc. supervisor) was that a manuscript should be written in the order that it is presented. The introduction informs what results should be highlighted as significant, and the discussion should contextualize said results within the previous findings highlighted in the introduction. I'm quite certain that I'm in the minority in following this advice - in my experience most people write the results (and methods) first. The only way that I could imagine not beginning with the intro is if you're writing a paper on a specific topic upon which you know the literature inside out - something I've never really had the luxury of doing as I've explored a variety of topics in several broad fields1.

Anyways, while I think that I've begun to hit a 'stride', the first few days of laying out my thoughts and writing were tough - more difficult than I remember writing being in the past. Though disconcerting, the reason is somewhat obvious - I went from a Ph.D. wherein I was writing all of the time (if not manuscripts, then fellowship applications, committee reports, a thesis, etc.), to a postdoc where I wrote very little. In fact, the writing style of my previous lab was rather non-standard, to say the least. For starters my PI would want to see the figures that you expected to put into any draft manuscript before he was even interested in seeing any text. Unfortunately, 1) making finalized, publishable figures takes a long time, and 2) it's difficult to determine how best to present genomic data without context - something I discussed in a previous post. Thus I spent far more time agonizing over the layout of graphs and plots than I did thinking about how to construct the story.

As I'm approaching yet another new field, beginning with a thorough literature review was crucial. Now it's just about working slowly through the results, seeing where adiditional analyses are needed, reworking my current draft figures to best illustrate my points, and writing it all down such that it's comprehensible by the reader.

Oh well, like I said, it's coming back steadily. Can you believe that I actually used to be among the few people I know who actually liked writing? Crazy hunh? Gotta get myself back in the zone! 

1I am trying to narrow, rather than broaden my expertise however, on the assumption that every scientist should find an appropriate niche.