My current work at D. E. Shaw Research (DESRES) centers on molecular dynamics (MD) and its application to biomolecular systems of scientific interest and pharmaceutical relevance. By modeling the motions of atoms within a molecular system, MD simulations can serve as a computational "microscope" onto phenomena that are difficult to observe experimentally, including the folding of proteins to their native three-dimensional structure, the structural changes that underlie protein function, and the interactions between two proteins or between a protein and a candidate drug molecule.
Unfortunately, these biochemical events often take place on time scales far beyond the reach of MD simulations on modern computers. One focus of our research is thus to develop new algorithms and computer architectures that make these time scales accessible by MD. We have created software that delivers an order of magnitude speedup over previously described codes for parallel MD runs on commodity computer clusters, and we are building a specialized, massively parallel machine intended to accelerate MD simulations by several additional orders of magnitude.
We aim to advance biochemistry and the process of drug development by applying these tools in-house. We are currently studying a variety of biomolecular systems, including kinases, membrane transport proteins, and ultrafast-folding proteins. We are also pursuing the development of improved computational chemistry methods and more accurate molecular mechanics force fields.
My research prior to joining
DESRES spanned two application areas (computational biology and
computer vision), but drew on common techniques (statistical modeling,
estimation, signal processing, and machine learning). To read more
about this work, click here.