It's more complicated than that
“Mad Hatter: “Why is a raven like a writing-desk?”
“Have you guessed the riddle yet?” the Hatter said, turning to Alice again.
“No, I give it up,” Alice replied: “What’s the answer?”
“I haven’t the slightest idea,” said the Hatter”
Lewis Carroll, Alice's Adventures in Wonderland
My first biology class was in the seventh grade. As is typical of biology classes, we spent a good deal of time memorizing things, as there is simply a lot that has been discovered, that needs to be taught. At the time, we learned the basics from the cell to the organelle to the macromolecule. Fast forward to college, I finally got the chance to work in a laboratory over the summer, and I looked into a dish of actual cells. They vaguely resembled what I saw in the textbooks, but messier. There was a lot more going on. "It's more complicated than that," I said to myself.
My interest in biology came from my long-time interest in curing cancer. When I joined the Cancer Biology PhD program at Stanford, I quickly learned one big reason why a tumor can outsmart every cancer researcher and oncologist on Earth. Heterogeneity. We're dealing with a clump of many different types of cells, all rapidly evolving down individual trajectories. One drug might wipe out some portion of a tumor, but if one cell out of a billion is resistant, that cell will divide over and over until it has replaced the tumor. "It's more complicated than that," I said to myself.
Upon starting the bespoke Cancer Biology program in 2011, there was a new paradigm that was taking biology by storm. It was called systems biology, and it was brought about in part by the -omics revolution. A series of novel technologies that would allow you to look at all of the genes at the same time (genomics), or RNA (transcriptomics), or proteins (proteomics), or protein-protein interactions (connectomics), and so on. These huge datasets could be analyzed by novel algorithms that would make sense out of all of it. I'll never forget what it was like to read genomics papers for the first time after being trained in reductionism. I thought that looking at things at the -omics level would bring clarity: clear-cut individual units or modules that brought about mechanism. I thought I would get to the final figure in the paper and finally understand how everything connects. More often that not, the figures would be really obscure visuals and big hairball networks. Very different than the clear-cut conclusions (gene X is a master regulator of process Y in cell type Z) I was used to pursuing. There were nuggets of insight in each study, but nothing near what I thought it would be. "It's more complicated than that," I said to myself.
My thesis lab was utilizing and developing a novel technology that would allow you to look at a ton of parameters, but at the single-cell level. The promise was that you could drop the word heterogeneity from your lexicon. You could finally look at a tumor and characterize every cell type, which of course would allow you to figure out what cocktails of drugs are needed for which cancer patient. I remember looking at this and seeing the fall of cancer. First leukemia, then lung cancer, then brain cancer, and so on. It was really exciting times. I first looked at immune system data using this technology, and it was nicely clear cut. You could see the forest rather than the individual trees. Then I finally looked at a cancer dataset. I was hoping that you'd be able to clearly see that different cell types correspond to different patients with different sensitivities to chemotherapy and radiotherapy, for example. Yes, there has been some success in this regard, but the data I looked at were surprisingly messy. The so-called heterogeneity was worse than I thought. "It's more complicated than that," I said to myself.
I went the bioinformatics route near the end of graduate school. I was using so-called nonlinear dimension reduction algorithms to quickly visualize the "map" of different cell types in these single-cell systems biology studies. One day, I finally questioned their efficacy. What I found was that the maps did a great job of characterizing the different "islands" of cell types. But fell flat in terms of the location of things within an island. By analogy, imagine looking at a map of Hawaii, and learning that on this map, the locations of the cities, beaches, and volcanoes per island were for the most part randomly generated. You could only take seriously island-level information: how many, how big, and a list of what's on each island. "It's more complicated than that," I said to myself.
What's the moral of the story? I'm a bit of an idealist. I sometimes find myself thinking that I'm nearing the end of the tunnel, and at the end is some deep intellectual enlightenment from which the cure for cancer (and the other major problems in the world) will just fall out. Sometimes I do experience deep insight (or perhaps the illusion thereof), and I go through a phase of complacency where I don't ask as many questions as I should. But then at some point I remember the hard truth. A truth that is perhaps the end of the tunnel itself: it's more complicated than that.
So that's the exercise that I give you here. The first thing a scientist does when beginning a study is to admit ignorance. The second thing a scientist must do through to the end of a study and beyond is to admit that "it's more complicated than that." To stimulate the process of asking questions and getting to the next level of understanding, even when you've learned a lot, simply say to yourself, whatever it is, "it's more complicated than that" and watch what it does to your mindset.