Skip to main content

Guangfeng Zhou, Ph.D.

  • Assistant Professor, Center for Advanced Therapeutics

Trained as a chemist and computational biophysicist, Guangfeng Zhou develops physics-based and AI-based computational methods to accelerate therapeutic discovery. His current program focuses on developing AI-based and physics-based computational approaches for drug discovery, including models that integrate physical and chemical principles, docking and binding-affinity prediction methods, and AI-accelerated virtual-screening platforms to identify and prioritize promising therapeutic leads.

Zhou earned a B.S. in Chemistry from the University of Science and Technology of China and a Ph.D. in Chemistry from Temple University, where his research used molecular dynamics simulations and kinetic network models to study protein folding and binding. He completed postdoctoral training at the Institute for Protein Design and Biochemistry Department at the University of Washington, where he developed computational tools for protein-ligand modeling, virtual screening, and therapeutic discovery before joining Wistar’s Center for Advanced Therapeutics.

View Publications

The Zhou Laboratory

gzhou@wistar.org

The Zhou Laboratory

The Zhou Lab develops computational methods and platforms that integrate physical and chemical principles with artificial intelligence to improve early-stage drug discovery. A major focus is the development of AI-based computational methods for structure-based drug discovery, with the goal of building models that generalize to new targets and chemical space beyond those represented in the training data. By improving the accuracy, efficiency, and generalizability of protein-ligand complex modeling and binding-affinity prediction, the lab aims to accelerate hit discovery and lead optimization.

The lab also develops physics-based computational methods for protein-ligand modeling, affinity prediction, and virtual screening, including tools within the Rosetta ecosystem. In parallel, the lab builds AI-accelerated virtual-screening platforms that enable more efficient exploration of vast chemical spaces and provide robust, deployable workflows for real-world drug-discovery campaigns. Through collaborations with experimental and translational teams, the Zhou Lab applies these approaches to therapeutic questions across cancer, immunology, infectious disease, and other biomedical areas. Importantly, experimental results from these collaborations provide valuable feedback for developing next-generation computational methods and improving the full virtual-screening workflow.

Available Positions

  • Learn about job opportunities at The Wistar Institute here.

Research

Integrating physics with AI for drug discovery

We are interested in developing AI models that incorporate physical and chemical principles to improve structure-based drug discovery. A major goal is to create models that generalize to new protein targets and chemical space beyond the data used to train them. By improving the accuracy, efficiency, and generalizability of protein-ligand complex modeling and binding affinity prediction, this work aims to accelerate hit discovery and lead optimization while reducing the cost and time required for early-stage drug discovery.

Physics-based protein-ligand modeling and AI-accelerated virtual screening

The lab develops physics-based computational methods for molecular docking, scoring, and virtual screening within the Rosetta ecosystem. These efforts focus on improving binding-pose prediction, binding-affinity estimation, receptor-flexibility modeling, and compound prioritization for experimental testing, with the goal of making structure-based drug discovery more accurate and scalable.

Building on these physics-based methods, we develop AI-accelerated virtual-screening platforms to explore vast chemical spaces with greater efficiency. These platforms combine structure-based modeling with machine learning to support the full virtual-screening workflow, from ultra-large library screening to compound ranking and experimental prioritization. The goal is to create robust, deployable tools that can be applied to real-world drug-discovery campaigns.

Collaborative therapeutic discovery

Our lab works closely with experimental and translational collaborators to move from therapeutic hypotheses to prioritized compounds for testing. These collaborations address therapeutic questions across cancer, immunology, infectious disease, and other biomedical areas. Current projects span diverse therapeutic modalities and target classes, including PROTACs and molecular glues for targeted protein degradation, GPCR agonists and antagonists, and ion-channel inhibitors. Through these projects, the lab applies AI-based and physics-based computational approaches for protein-ligand modeling and virtual screening to define tractable therapeutic targets, prioritize compounds, interpret mechanisms of ligand binding, and support hit-to-lead optimization. Importantly, experimental results from these collaborations provide valuable feedback that guides the development of next-generation computational methods and robust, deployable drug-discovery workflows.

Zhou Lab in the News

Selected Publications

An artificial intelligence accelerated virtual screening platform for drug discovery.

Zhou G., Rusnac D.-V., Park H., Canzani D., Nguyen H. M., Stewart L., Bush M. F., Nguyen P. T., Wulff H., Yarov-Yarovoy V., Zheng N., DiMaio F. “An artificial intelligence accelerated virtual screening platform for drug discovery.” Nature Communications 15, 7761 (2024). PubMed: https://pubmed.ncbi.nlm.nih.gov/39237523/

Force Field Optimization Guided by Small Molecule Crystal Lattice Data Enables Consistent Sub-Angstrom Protein-Ligand Docking.

Park H., Zhou G., Baek M., Baker D., DiMaio F. “Force Field Optimization Guided by Small Molecule Crystal Lattice Data Enables Consistent Sub-Angstrom Protein-Ligand Docking.” Journal of Chemical Theory and Computation 17(3), 2000–2010 (2021). PubMed: https://pubmed.ncbi.nlm.nih.gov/33577321/

Modeling protein-small molecule conformational ensembles with PLACER.

Anishchenko I., Kipnis Y., Kalvet I., Zhou G., Krishna R., Pellock S. J., Lauko A., Lee G. R., An L., Dauparas J., DiMaio F., Baker D. “Modeling protein-small molecule conformational ensembles with PLACER.” Proceedings of the National Academy of Sciences122(45), e2427161122 (2025). PubMed: https://pubmed.ncbi.nlm.nih.gov/41187076/

Automated identification of small molecules in cryoelectron microscopy data with density- and energy-guided evaluation.

Muenks A., Farrell D. P., Zhou G., DiMaio F. “Automated identification of small molecules in cryoelectron microscopy data with density- and energy-guided evaluation.” Structure 33(10), 1760–1770.e5 (2025). PubMed: https://pubmed.ncbi.nlm.nih.gov/40713967/

View Additional Publications