Decoding adaptive immunity by integrating interpretable deep learning and biophysical modeling to design immunotherapies

Maria Rodríguez Martínez leads a research program at the interface of computational immunology and deep learning. Her lab develops interpretable AI models to understand and engineer adaptive immune responses, with a focus on T and B cell biology. A strong emphasis is placed on transparency and explainability, ensuring that model predictions are not only accurate but also biologically meaningful. Current efforts center on modeling immune receptors by integrating protein sequence and structure with cellular information. To this aim, her lab combines coarse-grained simulations, molecular dynamics, and generative protein models to capture receptor flexibility and predict antigen recognition. These tools support the characterization of binding specificity, cross-reactivity, and off-target effects, enabling the design of safer immunotherapies and advancing our understanding of autoimmune diseases and other immune-related disorders. Her earlier work advanced the application of AI to cancer systems biology, including the development of deep learning algorithms for drug sensitivity prediction and the analysis of tumor heterogeneity using single-cell proteomics. She has also contributed to multi-scale models of immune cell differentiation and reinforcement learning frameworks for therapeutic compound design.

Methods

computational modeling ;  proteomics ;  statistical methods ;  transcriptomics  

Topics

algorithms ;  artificial intelligence ;  computational biology ;  immunology ;  machine learning  

Biography

Rodríguez Martínez is an Associate Professor in the Department of Biomedical Informatics and Data Science and a member of the Center for Systems and Engineering Immunology at Yale School of Medicine. Prior to joining Yale in 2024, she served as Technical Lead of Computational Systems Biology at IBM Research Europe, where she directed international consortia in precision oncology and immunotherapy. Trained as a physicist, she transitioned into computational biology during her postdoctoral work at the Weizmann Institute and Columbia University, focusing on systems biology, cancer modeling, and drug response prediction.