Vikram Sundar

By tackling difficult questions in computational structural biology, Vikram Sundar hopes to help scientists develop drugs more quickly, cheaply, and effectively.

He is currently a second-year graduate student in computational and systems biology at the Massachusetts Institute of Technology, and is working with Prof. Kevin Esvelt on applications of machine learning to continuous directed evolution for the design of targeted proteases.

Vikram was first introduced to the field of machine learning in computational biology as a Churchill Scholar at the University of Cambridge. There, he worked with Dr. Lucy Colwell on applying machine learning to predicting active ligands for protein targets, a crucial step in early-stage drug discovery. His most interesting result was a new method of training models to make predictions for targets without any prior screening data, allowing machine learning models to be used even with very little experimental data available. He also spent a year as an AI resident at Google, working on applications of machine learning to DNA-encoded library data for predicting protein/ligand binding.

More recently while at MIT, Vikram has become interested in the field of protein design, an important field with applications ranging from antibiotics to biomanufacturing. Beyond his current PhD work on protease design using directed evolution data, he has also worked on applications of machine learning to structure-based sequence design. He is currently interning this summer at Generate Biomedicines, a startup focused on using machine learning and high-throughput experimentation to enable protein design.

Outside of the lab, Vikram is passionate about diversity. As an undergraduate, he co-founded Gender Inclusivity in Math, an organization dedicated to reducing the gender gap in the Harvard math department. He is also an avid classical pianist and chamber musician and enjoys listening to and playing works by Beethoven, Chopin, Liszt, and Rachmaninoff.