Sushmita Roy is an Assistant Professor in the Biostatistics and Medical Informatics Department at the University of Wisconsin, Madison. She received her B.S. in Computer Engineering at the University of Pune, India. She received her Ph.D. in Computer Science in 2009 from the University of New Mexico and did a post-doctorate at the Broad Institute of MIT and Harvard. She is a recipient of the 2014 Alfred P. Sloan Foundation Fellowship and an NSF CAREER award.
It’s not just genes that make us who we are. The way that genes are regulated and turned on or off (also called epigenetics), rather than the DNA sequence of the genes themselves, is also very important. Dr. Roy’s lab focuses on the development and application of statistical computational methods to identify the gene regulation networks which drive cellular functions by integrating different types of genome-wide datasets. In her seminar at the Buck Institute, she first explained why mammalian gene regulation is complex:
- There are thousands of transcription factors with unknown binding specificity
- Gene expression is regulated by the interplay of transcription factors and chromatin (chromatin is the combination or complex of DNA and proteins that make up the contents of the nucleus of a cell)
- Regulatory DNA elements are not necessarily next to the gene
- 3D organization of the genome also plays a role in gene expression.
She also introduced three projects from her lab:
- Gene regulatory network which controls host response to different strains of influenza infections
- Examining chromatin state model to help identify epigenetic barriers in cellular reprogramming
- Predicting target genes of candidate enhancers (regulatory sequence elements).
SAGE sat down with Dr. Roy to ask a few more questions…
Q: Could you summarize your research to the general public using non-scientific words?
SR: I am interested in understanding how cells know what to do when. So basically more complicated organisms have different types of cells, and each cell has different types of functions. One of the ways that cells are able to do a particular specialized function is by expressing the right type of genes or the right set of genes. I am interested in developing computational methods to try to understand the molecular circuitry of cells that determine what genes must be expressed when and where. Specifically, we want to know: what is the underlying gene network in a particular cell type? How does the network change between different cell types or different environmental conditions?
Q: Can you tell us a little more about how chromatin state plays a role in gene expression?
SR: When I say the chromatin state, I mean that you should think of DNA as not just a string of letters. DNA is wrapped around histone proteins, and this is how the DNA is packaged inside the nucleus of a cell. These histones get modified bio-chemically, which in turn influences the accessibility of the associated DNA to other important proteins that activate genes (e.g. transcription factor proteins). In particular, some modifications make the DNA more conducive to activate expression while some modifications make the DNA less open and conducive to expression. Expression of a gene in turn is controlled by the set of transcription factors that are bound to the gene’s promoter. We are still figuring out the interplay between chromatin state and transcription factors. As more datasets from multiple cell types become available, we hope to get a better understanding of the relative importance of chromatin state, chromatin modifying enzymes and transcription factors in specifying gene expression levels.
One would expect that when chromatin state changes, the associated gene expression levels also change. However we see that this is only partially true.
Q: What is the main challenge in your field right now? We know that the machine learning model involves using known data to make predictions of the unknown. Can you use your model to predict?
SR: A major challenge in the network community is the lack of good gold standards of a “correct” regulatory network that is large enough to get realistic estimates of how a method for network reconstruction works. Interpretation of results is also a challenge, again, due to the large number of unknowns. We can use our models to predict expression in a new condition provided we can measure the activity of some of the components of the network. The big picture would be to try to predict how a cell behaves in new perturbation – something that you’ve not measured.
Q: Which research project is the one you’re most excited about right now?
SR: All three of them are very exciting projects that ask pieces of a bigger question of building a predictive model of a cell’s expression profile. I’m very interested in delving deeper into the chromatin state and its connection to mRNA levels. That link is not very well understood, and I want to have a bigger picture of what it is. You can think of chromatin modifying enzymes also as potential regulators. I want to try to understand the connections among these different chromatin modifications and how that affects the structure and function of a regulatory network driving a particular process or a particular phenotype. That certainly is something that I want to spend more time on and understand more. Ultimately, I would like to gain a better understanding of transcriptional regulation as a function of the chromatin state, transcription factor occupancy, signaling networks and the organization of the genome and how network change impacts complex phenotypes.
Q: Are you planning to conduct studies that collaborate with aging research groups?
SR: Certainly if there is an area where I can add my expertise to. I would be certainly enthusiastic about working on aging, which I think is a very important problem. It would be very interesting to study the connection between diet and aging as well as aging and different diseases. So if approaches like mine can be used to try to address such problems, I would be enthusiastic to collaborate.
Q: Can you give postdocs advice for how to succeed in an academic career?
SR: You have to figure out what you want to do, what’s exciting to you and let that drive you. You need to ask yourself if you like research and if you’re driven by it. It’s also important to acquire skills that enable one to collaborate with a diverse set of scientists from different disciplines including from computational and biological areas.
I think having a good postdoc experience is also important. For me, my experience was great because I was working on what I wanted to do, and I was fortunate enough to meet people who were willing to support me. I am originally from India, and I was fortunate to have the opportunity to pursue higher education in the US. I always wanted to do something in medicine and biology, but I was a computer engineer. I really wanted to come back to biology, and I was fortunate to meet my PhD mentors in New Mexico to support my interdisciplinary focus of computer science and biology. For my postdoc at the Broad Institute, I dived deeper into computational biology, and I was surrounded by computational scientists as well as experimentalists who were working in close collaboration with the computational scientists to address important questions in biology. That was a very useful and nurturing experience for me.
Q: As an international research scientist, what kind of difficulties or challenges have you faced, and how did you overcome them?
SR: When I came here, I came with uncertainties about funding, so that was bit of a worry initially. Making sure I had some way to support my education was important and once that was taken care of, I really enjoyed being here. I was also fortunate to have great advisors to support my PhD efforts. Overall, I’ve been around people who’ve been very supportive and encouraging mentors. I think I am pretty fortunate in having these kinds of people to help me understand how to be a scientist and researcher.
Q: What is the academic environment in India?
SR: It’s getting better but it’s not nearly as great as it is over here. There are some institutes with new PIs over there with a lot of energy, who have funded labs doing great work. But in general, we don’t have the resources and the infrastructure like the US. India is behind, but it’s getting better and better. Certainly having more funding and more support would help things to improve. The government is supporting, but I think there needs to be more.
For more on Dr. Roy’s work, check out the Roy lab website.