Vinay Saji
Mathew

Hello! I am a Ph.D. student in the Department of Industrial Engineering at Penn State. My research focuses on algorithms that accelerate protein structure prediction while preserving physical validity. Structure prediction sits inside drug design and virtual screening pipelines and is often what limits how far they can be used. The idea is to match the quality of the large teacher model in fewer diffusion steps. This work is a collaborative effort with the Lai Lab at Cornell University. In parallel, I contribute to a DOE Genesis Mission project developing neurosymbolic methods for manufacturing. The neural part learns from massive amounts of live data and helps shrink the space of possible decisions, working in unison with a traditional solver. Over the years, I've been fortunate to have generous support from the Epigenomics Facility at Cornell University, the Institute for Computational and Data Sciences at Penn State, and the Department of Energy, with computing support from the NSF (Award #CIS260116). We open-source all of our work here.