Developing machine learning systems that reason about molecules, representations, and chemical environments.
We develop machine learning methods for understanding molecular structure and function. Our research focuses on graph neural networks, generative modeling, representation learning, and property prediction across chemical spaces.
A major goal is designing unified models capable of transferring chemical intuition across diverse molecular tasks.
We introduce a machine learning framework for predicting molecular properties across chemical mixtures using graph neural networks.
Modeling complex multi-component chemical mixtures, odor perception, and formulated products.
Most real-world chemistry involves complex mixtures rather than isolated pure molecules. We build machine learning algorithms to model mixture interactions, olfactory perception, solvent dynamics, and metabolic networks.
We introduce a machine learning framework for predicting molecular properties across chemical mixtures using graph neural networks.
Integrating AI planning with automated robotics for high-throughput chemical experimentation and discovery loops.
We combine Bayesian optimization, robotic automation, and generative AI to create closed-loop self-driving laboratories capable of discovering novel functional molecules and materials autonomously.
