February 13, 2025 Webinar
Advancing Programmable Human Digital Twins for Substance Use Disorder Drug Discovery
Lei Xie, PhD
Abstract:
In the quest for effective, safe, and personalized treatment for Substance Use Disorder, traditional one-drug-one-gene approach often falls short due to its inability to fully capture the complexity of human pathology and fill in the translational gap from early drug discovery to clinic. We propose a paradigm shift towards a system pharmacology framework, leveraging the rich data from patient and perturbation omics profiling. This approach aims to transform disease states into healthy ones by identifying key biomarkers and system-wide modulators through multi-scale predictive modeling of drug actions in the human body. However, there are challenges ranging from the integration of diverse omics data types to the exploration of uncharted chemical, target, and functional genomics spaces. To address these challenges, we leverage advanced AI techniques for their power in data-driven feature extraction, harmonizing heterogeneous data sets, and multi-level modeling of complex biological systems. Our recent efforts include 1) harnessing the vast volumes of unlabeled data to illuminate the dark corners of chemical and biological knowledge, 2) cell type-specific phenotype compound screening across biological levels, and 3) Transfer learning to bridge disease models and human physiology. Put together, the AI-powered systems pharmacology approach has been successfully applied to personalized Alzheimer's disease drug repurposing and Opioid Use Disorder drug discovery.
Short Bio:
Dr. Lei Xie is currently a professor in Computer Science at Hunter College, and Ph.D. program at Computer Science, Biochemistry, and Biology at the Graduate Center, The City University of New York. He is also an Adjunct Professor in Neuroscience at Weill Cornell Medicine, Cornell University, and Chief Scientific Officer of Dark Matter Therapeutics, Inc. His research focuses on developing new methods in artificial intelligence, machine learning, systems biology, and biophysics for multi-scale modeling of drug actions and causal genotype-phenotype associations, and applying them to drug repurposing, drug discovery and precision medicine. From 2001 to 2011, he was a principle scientist at San Diego Supercomputer Center (SDSC), research scientist in pharmaceutical company Hoffmann-La Roche and biotechnology start-up Eidogen. He was trained in Computational Biology and Biophysics as a postdoctoral fellow at Columbia University and Howard Hughes Medical Institute from 2000 to 2001. He obtained his Ph.D. in Medicinal Chemistry from Rutgers University, and B.S. in Polymer Physics from University of Science and Technology of China.