Reflecting on my first major scientific publication.
Today, my paper “ARID1A loss impairs enhancer-mediated gene regulation and drives colon cancer in mice” was published in Nature Genetics. A second paper from my lab “SMARCB1-mediated SWI/SNF complex function is essential for enhancer regulation” was published simultaneously, also in Nature Genetics. These papers are the culmination of several years of hard work with collaborators across Harvard Medical School and other institutions. Together, they establish an epigenetic mechanism of tumor formation that is fundamental to the biology of a large fraction of human cancers, and we hope they will translate to improved patient care in the near future.
My project required me to breed, inject, dissect, and analyze the intestines of several hundred mice. I’m grateful to Boris Wilson for introducing me to mouse work gently enough for me to overcome my psychological and physical revulsions, and to Adrianna San Roman for sharing with me her perfected techniques and invented tools for flushing poop out of pieces of mouse colon. Three generations of impressive lab technicians — Jeff Haswell, Haley Manchester, Weishan Wang, Kelly Sullivan, Ev Tzvetkov, Tam Pham, Emma Troisi — helped keep my mice (mostly) alive and (mostly) accounted for. Expert gastrointestinal oncologists, Dr. Ramesh Shivdasani and Dr. Tony Agoston, spent valuable portions of their time giving me individualized lessons in histopathology, which I immediately put to good use on the microscope, distinguishing tumor from normal and abnormal-but-not-tumor tissue, always losing track of time, and being too disoriented to focus my eyes or attention back to the macroscopic world.
In addition to the mouse work, I learned to perform a whole host of biochemical experiments involving excruciatingly detailed multi-week long protocols, microliter (sometimes nanoliter) quantities of fluids, and multimillion dollar machines capable of handling them. I learned from Xiaofeng Wang, an extraordinarily skilled scientist who guided me through protocols and who I concluded had “magic hands” after he successfully performed an experiment after I had failed five times — I never did figure out which step(s?) I had been doing differently from him. By the end, we had gotten ChIP-Seq, RNA-Seq, Exome-Seq, and every other relevant, thinkable experiment to work. No technique, no matter how complex, was beyond our capability.
With data came bioinformatics which, for my project, were everything but standard. I had ventured into unexplored biological terrain: I wanted to analyze changes not just at one gene, but at all genes, and to focus not on annotated regions of the genome, but on regulatory elements in regions of the genome often dismissed as “Junk DNA”. My A-grade in the Harvard Genomic Data Manipulation course did not qualify me for the task, and so I turned for help to Burak Alver, an MIT high-energy-physicist turned Harvard bioinformatician. After months of tortuous email communication, I walked down the block to his office and we sat down together to try to processes my vastly complex datasets and answer the equally complex molecular questions in my head. We began to appreciate the value of true collaboration as we together devised the cleverest ways to turn our data into interpretations and then to figures — tweaking code over and over until the visuals we had created were satisfyingly “pretty”.
Writing the manuscript was an entirely different challenge. In this, I was completely alone. I re-arranged words in sentences and sentences in paragraphs, trying to achieve the elusive “flow” while ensuring a priority order of accuracy > clarity > style. Like with every other writing-phase of graduate school—the dissertation committee reports and NIH grant proposal — I worked so late into the night that the 6am bus passing my apartment became my signal to give up and crawl into bed. I became convinced that slipping in and out of consciousness in the middle of the night actually worked for me as a writing technique. It certainly seemed to be the only way I got words on a page.
Regarding the final publication — I think I ultimately failed in my mission to make it easily readable. However, I am more forgiving of myself now that I have been at any point over the last three years. My stress and anxiety levels have been high, and I have raised them higher by considering being stressed and anxious to be personal failures in and of themselves. Only now am I realizing how truly HARD all of this has been. It wasn’t just that the mouse and laboratory work were challenging, requiring levels of patience and technical precision that were never natural for me, it was also the nature of science itself. Each experiment carried a risk that I would get results I could not explain or that I would get no results at all — our models could be wrong; our tools advanced, but not advanced enough. It was also, in retrospect, a fool’s errand to describe a project this complex in a way that was easily readable to those outside the field, and to have it fit within the journal’s 1,500-word limit (my methods section ended up being over twice as long as the text!) The writing-phase was also never just about putting words on a page. It was the period of time that required some of the hardest thinking that I have ever done in my life. It’s no wonder it was often difficult to settle on the right words; I was, in effect, inventing language to describe molecular phenomena I was observing for the first time — phenomena that may never have been conceived of by anyone else! Given all of this, it was quite audacious for me, starting out as a 21-year old graduate student, to try to do any of this at all. To believe I could reach the frontier of human knowledge and push it even further. To think I could take on not just any project, but one that was hugely consequential for human disease. And not just any disease, but cancer. The emperor of them all.
My advisor, Charlie Roberts, often told me I was being too hard on myself. His words fell on deaf ears — I was pre-occupied either pretending I was fine, or I had already burst into tears and was frustrated with myself for allowing that to happen. “Hard on myself” sounded like yet another personal failure to add to the list. In reality, Charlie had believed in me from the very beginning. He had allowed me to develop my own project, move forward with the experiments I thought would yield the most interesting results, and had never questioned my (highly questionable) writing techniques. I had never directly asked him for independence; as an imposter-syndrome-afflicted early-year graduate student, I didn’t even necessarily want it. Looking back, I cannot imagine not having had it.
Charlie left Harvard in the middle of my PhD to become the Cancer Center Director of St. Jude Children’s Research Hospital in Memphis, Tennessee. While I recognized it was a tremendous honor and opportunity for him, I was upset by where it left me. I was in my third year of graduate school, with more experiments underway than I could keep count of, and I had just been awarded an individual NIH grant for my project. Now, all of the sudden, I was told my options were to either move to Tennessee or to find a new lab to join at Harvard. Both sounded terrible.
Thankfully, I found a third. The Chair of the department, Stuart Orkin, allowed me and a few others to occupy space in his laboratory to finish the projects we had started under Charlie. Stu was a world-renowned hematologist-oncologist who had trained Charlie and many other now-senior scientists at Harvard. I had presented several times at the departmental meeting he hosted, but had never directly spoken with him. I went ahead and declared him my new advisor. I was nervous before our first meeting, but I needn’t have worried. Stu was wonderful. A “man of few words”, the few words meant everything. If Stu thought everything would turn out okay, it would turn out okay.
And now it has.
I am so incredibly grateful.
I am so incredibly proud.
