Keeping Busy

Sometimes, despite all my best planning, I run out of tasks that need to be done. This seems to happen to me most often in the experimental phases of a project, where I first need to do experiment A, and see what A looks like, before deciding whether experiment B or C is the most appropriate direct follow up.

There are, nevertheless, things I can do during this time while I’m waiting for A to finish (other than wasting time on Facebook or the like).


Ultimately, the experiments I’m doing will be written up as a paper, and I’ve found it very productive to start writing the paper, even though I don’t have all (or any) of the results, and the methods are still under development. At one point, I even put up a calendar above my desk with a goal of just writing for at least half an hour a day, every day, inspired by none other than Jerry Seinfeld.

Do you know what general questions your experiment is going to answer? Cover that in the introduction! There’s also plenty of room here to go through the past literature and start putting together references that you know will be relevant. Some of this background will also be useful in the discussion section.

Most surprisingly, though, you can also start writing your results. I’ve got a couple drafts sitting around on my computer that have things like:

We identified NNNN clusters in SPECIES, and NNNN clusters in OTHERSPECIES. We then manually grouped these clusters into NNNNN coherent super-clusters.

We found that within each of these super-clusters, genes DID SOMETHING THAT I’D NEED TO SEE THE ACTUAL DATA FOR.

You can’t imagine every contingency, but there’s lots of places where you can start laying out the experiments you plan to do, and imagine the results they’re going to give you. Whether those results will be numerical, binary, or more qualitative, I just put in placeholder text that will be obvious to catch later (for example, by egrep '[A-Z]{4,} manuscript.tex to find chunks of all caps).

This is also a nice time to write because you will have to more clearly identify your personal biases. If you haven’t done any of the experiments yet, you shouldn’t already know what the answer is. Therefore, if you’re writing up the results as if you do, then perhaps something is off. It’s also nice to be able to sit back and think, “If someone else claimed that this happens in an experiment, what kind of controls would I look for them to have done to convince me”.


After I’ve written in a journal-article-like form what I’m planning on doing, it’s often possible to start implementing some of the analyses in code. If you don’t have a relatively simple set of commands to run that will generate your figures, how will your reviewers ever check your code?

What, your reviewers aren’t checking your code? Actually, this is not yet something reviewers actually do, but maybe they should. The first step in saying it’s reproducible science, after all, is a fast and easy reproduction (with the same raw data).

You might say, at this point, “But Peter, I don’t have any data”. That may be true, but you should have some sense of what your data will look like. Simulate it however you like, and at least check that your analysis pipeline is able to pull out what you put into the simulation.


I originally conceived of this post when it looked like I was going to have some down time waiting for reagents to come in, but the magic of fast shipping killed that leisurely week. Still, it’s always good to get your brand out there, right?

Do I follow all the above advice? Not as much as I should. But I try, and it’s always possible to keep busy!

Leave a Reply

Your email address will not be published. Required fields are marked *