Sandeep is a sixth year PhD student in Dmitri Petrov’s lab and Biology’s Ecology and Evolution Department. In 2010, he received his B.A in Biology and Computer Science from Washington University in St. Louis, where he completed his honors thesis with Dr. Justin Fay and studied the evolution of gene expression regulation in yeast. His research focuses on studying adaptive evolution, using both experiments and simulations.
This content has been transcribed from an interview that took place on Stanford campus Wednesday, January 13, 2015 with CEHG’s Director of Programs, Cody Montana Sam and Communications Manager, Katie M. Kanagawa.
Can you start by telling us a bit about yourself and your science background?
My work is mostly on experimental evolution and evolutionary theory, studying the predictability of evolution using model systems to try and characterize adaptation, particularly how much adaptation occurs, what kinds of mutations they are, how good they are for the organism, and how they interact with different environments.
As for me, I was born in Syracuse, New York and moved to a D.C. suburb in Maryland when I was two. My dad is a chemist and my mom is a programmer, so science fluency was very much part of the way I was raised, whether I was reading National Geographic books, watching NOVA programs, or doing various scientific extracurricular activities. I grew up in a place with the Smithsonian Institution which has, as far as I’ve seen, the best museums on the planet, so I spent all my summers there. I was also lucky enough to have a lot of great biotech institutes in the Maryland area, especially the University of Maryland’s satellite campus which was close to my house and had a bunch of labs. They had a high school research program that I started attending in junior year and I got hooked. I really didn’t consider anything else as a career after that.
How did you decide on Petrov’s Lab? It seems like a natural fit.
You know, my trajectory through Biology has been kind of weird. The first lab I worked in during high school was a biochemistry lab, where we were studying the protease subtilisin, just characterizing the active site and mutations and playing with the energetics of it. The following summer was another lab with a computational biochemist who was studying protein folding, so I did some coding with him. Then I spent two summers at a prostate cancer lab, doing immunofluorescence and western blots, PCR, all kinds of basic molecular things. Then my undergrad thesis was doing gene regulation evolution with Justin Fay; that was my first serious, long-term project and I actually managed to get a peer-reviewed paper out of it.
I’ve always been interested in the mechanistic, theoretical underpinnings of things, and after my work with Justin, I wanted to get more into how exactly we know that something is evolving.
Comparative genomics is kind of messy, just because it is all about inferring evolution rather than observing it, so getting into questions like how you know something is really adapting and what is the best experiment we can do to say that this is actually what is going on drove me to experimental evolution. So I joined Dmitri’s lab after doing a rotation with Hunter [Fraser] and Dmitri and I’ve been with Dmitri ever since.
What has it been like for you, working with Dmitri and Hunter?
Working with both of them has been a lot of fun. They do very different things. I found myself drawn to more fundamental evolution questions as opposed to the applied gene expression work Hunter does, which is why I joined Dmitri’s lab. It’s been a blast! I’ve been in on a number of different projects doing very different things.
I have one purely theoretical project where we just simulate evolution, trying to see how predictable it is. Then I’m on this very large collaboration between the Petrov group, Gavin Sherlock’s group, and Daniel Fisher’s group doing experimental evolution in yeast, where we have barcoded yeast cells with hundreds of thousands of unique DNA tags. This gives us the resolution to start asking what individual mutations are doing in a large evolving population. When you have a yeast population, even in a flask, there are billions of cells. If you want to track a mutation that’s potentially adapted, by the time you can see it through whole genome sequencing or other methods, it has already reached millions and millions of cells. That’s kind of too late to really understand what’s going on and you’re going to miss a lot of the big picture. So one of the former postdocs in Gavin’s lab came up with this idea of tagging individual yeasts with DNA barcodes, just random sequences, and sequencing that to track the frequencies of subpopulations of cells over time. So if one cell gets an adapted mutation, then the barcode that’s associated with it will start rising in frequency as that group of cells starts expanding in the population. You can detect it very early on and get a better picture of how much adaptation is going on, you can quantify how fit they are and how fast it happens, and how the long-term dynamics play out. That paper was published last year.
My project then became to characterize what exactly the mutations were. Now we have a set of hundreds of clones where we know what the mutation is and how much of an advantage it gives to the cell. We can start getting statistics on that process, and get a fairly comprehensive understanding of how the population is adapting at the molecular level.
Do you have a plan for where you are going next, after you finish your degree in six months?
My dream research goal is to be able to start from a very mechanistic basis for some arbitrary mutation and use a computational model to infer what it does to the function of a protein, what that does to its protein complex, and how these effects propagate to the whole cell, groups of interacting cells, the whole population, and finally the whole ecological system. From there, we can start trying to infer how this mutation changes the fitness of the organism, and how that changes as a function of the environment or other genetic backgrounds. It would really be ideal if we could model the effect of a mutation from a systems-level perspective, starting from an individual mutation to what it does to a population and how that changes the fitness of the organism. There is currently no system where we have close to enough knowledge to pull that off but we need to develop techniques to be able to predict what particular mutations will happen if we give particular organisms an evolutionary challenge. That’s important generally, not just for yeast evolution but for things like cancers, drug resistance evolution, and other industrial applications where evolution is useful because you want to know what will happen in the future given the current state. That is where I want to end up going.