Olexandr Isayev

Olexandr Isayev has developed neural network-based methods for molecular property prediction and virtual screening at unprecedented scale, enabling AI-driven evaluation of billions of compounds against cancer drug targets in hours rather than the weeks required by traditional docking methods. His ANI neural network potentials achieve quantum mechanical accuracy at molecular mechanics speed for drug-relevant calculations. He has pioneered active learning approaches that efficiently explore chemical space by iteratively selecting the most informative compounds for experimental testing, dramatically reducing the number of experiments needed to identify potent cancer drug leads. His uncertainty quantification methods enable reliable prioritization of virtual screening hits for synthesis and testing.