My research interests lie at the intersection of biology, physics, and machine learning, with a focus on understanding and predicting the behavior of biological systems. The behavior of living organisms at different scales requires a deep understanding of the collective behavior of molecular machines, primarily proteins. Jacques Monod, a visionary who sought to put biology on firm mathematical grounds, viewed proteins as the ultimate teleonomic agents that animate all forms of life through their purposeful activities. They can recognize and bind with "elective discrimination" to other molecules. In most situations, they exercise their faculties by spontaneously folding onto three-dimensional geometrical structures. John von Neumann's Theory of Self-Reproducing Automata provides a general framework for thinking about the complex behavior of living organisms as machines that can self-replicate, evolve, and produce inheritable traits, with proteins serving as the primary hardware and one-dimensional chains of nucleic acids storing the software.

However, accurately simulating the collective behavior of molecular machines remains a daunting challenge, with the combinatorial explosion of the number of molecular machines and their conformations, as well as the number of possible interactions, making experimental methods alone insufficient. Recent breakthroughs in the field, such as AlphaFold2 and our OpenFold, offer a promising path forward. These models use machine learning to predict the structure of proteins with unprecedented accuracy and have the ability to transform our understanding of how we think about cells and organisms.

Inspired by the success of foundation models trained on diverse data that can perform a range of downstream tasks, my vision is to create a unified model that can generalize across different domains and scales of biology, turning biology into a predictive science. By developing a deeper understanding of the principles underlying life's complexity, we can unlock new avenues for designing and engineering novel biological systems, with applications ranging from drug discovery to sustainable agriculture to biotechnology.

In collaboration with Mohammed AlQuraishi, and building upon OpenFold, I am leading multiple research projects aimed at advancing our understanding of protein and RNA structure prediction, ligand binding, and machine learning for molecular modeling.