On what I love about science
I have written negative posts about science, consensus, knowledge, which may make it seem like I am resentful of my chosen profession. I am not, I love science.
My day-to-day work is a lot like a chef. Multi-tasking constantly, laying out and rearranging timelines in my head, working with my hands, seeing my labor produce results in real time. The ability to multitask is what makes a good scientist, and if I have one talent it’s that. I get in to lab first, around 7 AM, and think “If I set up this PCR now I’ll have 1 hour of downtime to passage my cultures, when it’s done I’ll have 15 minutes for a fast gel, then 10 minutes of active work setting up the next reactions, while that goes I can start staining my cells and pour some agar plates for a transformation I’ll do towards the end of the day, find the DNA stocks I need and make growth media for an experiment I’ll set up tomorrow, and so on…”
When I cook the rice timer clicks, the roasting vegetables sizzle, and the beef hits 120F within 30 seconds of each other.
It's intense work if you’re doing it right, I eat lunch in 15 minutes, my 1 PM coffee is down in 2, I am rarely sitting down for more than 5 minutes at a time. There’s little moments in the day where you think “fuck, should I set up this next thing or am I doing too much” and you have to push through. When I present in lab meeting people are shocked at how much I did in a month. The little moments of pushing though add up. Good fundamentals and pruning down every protocol to it’s bare essentials makes the work much faster, which means you can handle more at once.
You don’t have to purify DNA for gibson assembly, you can use GXL polymerase at 10 seconds/kb instead of 60 with 25% of the recommended polymerase, no need to recover an ampicillin transformation, no need to recover yeast transformations for auxotrophic markers, 15 minutes is plenty for most gibsons, staining antibodies are done in 10 minutes not 30, lithium/boric acid let you run a gel 3x as fast as tris-acetate, PI staining can just be added into the last step of flow cytometry prep with no washing. A key feature of what sets someone apart is taking the advice of their mentors, deconstructing and finding the bits that might be based on legacy fake assumptions.
Planning the work requires imagination, rigor, vision, and creativity. In biology you can’t ever prove anything. All you can do is eliminate alternatives. That means for each experiment you must think about all the possible outcomes, the trivial explanations, how you’ll separate the question you’re asking from the noise of biology. Without the right controls your results are meaningless, and you can get dragged into months of a wild goose chase, getting caught in an ever more complicated mental construct which misunderstands what you’re studying. Sometime, if a tiny little thing in your work environment changes it’ll break your work for weeks to months at a time, that sucks. It’s happened to me a couple times, makes you feel crazy, nearly impossible to figure out.
In the best case you’re trying to build a mental model for what you’re studying. It’s a disorganized spiderweb with the thing in the middle. Close to it is the knowledge that is most relevant and sturdy. DNA codes for genes, RNA is turned into protein, etc. Then is the next level of slightly less certain information, cells have states, the mitochondria does this. Then the next level, neutral drift is a important force in evolution, T-Cells get exhausted, GPCRs signal in a biased way. This should be considered little more than noise. Maybe its true, maybe its not. In your mental model, if this is true, how does it pull on this thing, and how does that pull on this thing? If it isn’t, how does that pull on those?
I am 3.8 years into my PhD. I am finally at the stage where I am closing out a couple “safe” projects based on my lab’s technology (in which I found a lot more quirks than was realized). Those projects are over the hump though, where I know they will produce publishable results and a couple of very tidy papers which show off some nice technical work. They are not mind blowing but show good execution and vision. A third one is almost there, but not as tidy.
On my last project I took a risk. The project originated from the flagellum rabbit hole I wrote about here. It is based on a take that evolutionary biology has missed a big, fundamental property of gene evolution. If that’s right, it could have implications in human health, manufacturing, and the fundamental way that life works. If it’s not, than: oops.
I started it relatively late in my PhD, 2.5 years in, and was told it would be impossible by my PI, and for various technical reasons by most people around me. I wouldn’t have tried it if I hadn’t got really lucky and won a fellowship that gave me the circumstances to be comfortable enough to not rush. I have overcome the technical hurdles that were guessed by the naysayers, found different ones, dealt with most of those, and the system I’ve built is good. I know the last few things I must do get it closer to excellent.
The system will let me ask some questions that are largely unprecedented. I’m running the first serious experiments with it while I make the last few improvements to the technology, and I have long since left the shores of relying on the literature behind. For now it’s a little secret, as if it works how I think it might then I will patent it and see if I might be able to make some money off of it. If it’s not, than: oops.
Being in this open experimental space is the best part of science. It took a full year of grinding with no guarantee of payoff, not much support, to get out to a place that no one has ever been. Many scientists take an iterative approach and never reach this space in a way other than not knowing the result of the most immediate level. I’m on the frontier with a wide open decision space in front of me and no one on the planet who knows what to do next better than me.
I am running experiments with it that are now spitting out the first data. I just passed some major hurdles and am pulling out new things that no one has ever seen. I’m corresponding with scientists I admire. I have some ideas to test.