Regina Barzilay

Regina Barzilay has developed graph neural network and generative AI approaches for drug discovery that have demonstrated the ability to identify novel chemical matter with desired anti-cancer properties from vast chemical spaces. Her laboratory's AI-driven discovery of halicin, a novel antibiotic identified through machine learning screening, provided proof-of-concept that AI can discover structurally novel bioactive compounds missed by traditional screening. She has applied similar approaches to cancer drug discovery, using generative models to design molecules optimized for potency against specific oncology targets while maintaining drug-like properties. Her work on AI-guided retrosynthetic planning enables practical chemical synthesis of AI-designed cancer drug candidates.