Life Sciences & Health

AI in the Life Sciences and Drug Discovery

A nucleus of LSH is the Department of Life Science Informatics (LSI) at the b-it with a strong track record in developing computational methods for pharmaceutical research, data analytics, and drug discovery applications. Methodological focal points include methods to identify new active compounds in virtual databases, chemical language models for generative design of highly potent compounds with desired biological activity, and chemically intuitive concepts for rationalizing predictions of Machine Learning models. Practical drug discovery efforts currently concentrate on the development of new inhibitors of understudied human protein kinases for cancer treatment and immunology.

AI in Medicine and Healthcare

AI in medicine demands models that are trustworthy, explainable, and privacy-preserving. At Lamarr, we develop theoretically grounded methods that integrate causality for counterfactual reasoning and explainability, including in foundation models. Together with clinical partners, we design federated and decentralized approaches enabling secure multi-center learning on sensitive health data. Our goal is to translate cutting-edge AI into reliable tools that directly improve patient care through early detection and robust decision support.

Explainable AI in Interdisciplinary Research

The impact of AI in the interdisciplinary research environments life sciences, drug discovery, and medicine largely depends on the use of computational results for experimental design and practical applications. Pooling expertise across research areas, the Lamarr Institute advances XAI in Life Sciences and Health to rationalize predictions and translate them into experimentally testable hypotheses for causal reasoning. Lamarr researchers hereby develop AI models that are transparent and understandable to an interdisciplinary audience.