Our lab's core mission is to decode the "language of chemicals." We operate on the principle that a molecule's structure contains vast, predictive information about its biological function. Our goal is to translate that structural information into a deeper understanding of health and disease.
What differentiates our approach is the integration of chemical and biological data. Instead of analyzing chemical structures in isolation, we build models that represent molecules within a combined "bioactivity-chemical" space. This enables us to contextualize a chemical's properties in relation to its known biological effects, resulting in models with superior predictive power and real-world biological relevance.
To build these sophisticated models, we enrich them with decades of established biological knowledge, integrating multi-omics data, including genes, mutations, and pathways. We develop and apply novel computational methods rooted in artificial intelligence, multimodal data integration, and emerging Large Language Model (LLM) architectures.
Developing predictive and explainable AI models to solve critical biomedical challenges, from understanding disease mechanisms to identifying novel diagnostic signatures.
Using our integrated models to understand the molecular drivers of aging and to discover novel geroprotectors-compounds capable of extending healthspan.
Creating new computational frameworks to predict a molecule's function, (e.g., carcinogenicity, therapeutic potential, or odor) directly from its structure, accelerating the discovery of new drugs and functional chemicals.