Líneas de investigación

• Artificial intelligence immunodiagnostics

  We develop artificial intelligence immunodiagnostics to predict patient outcomes from immune cell states in blood. By combining single-cell transcriptomics with domain-aware machine learning, we identify conserved gene programs (Kunes & Walle Nat Biotech 2024) and use artificial intelligence to link immune activity to therapy response. This approach moves beyond simple blood counts to capture intra-cell type heterogeneity, uncovering mechanistic insights and new therapeutic targets. Ultimately, we aim to use the immune system as a ubiquitous biosensor for cancer, infection, and other health exposures. 

• Understanding human systemic cancer immunosurveillance and escape

  We aim to understand how human immunosurveillance shapes the earliest progenitors of cancer (Burdziak et al Science 2023). Most knowledge comes from mice or invasive tumors, but little is known about premalignant immune control in humans. Using Lynch syndrome as a model with defined antigens and paired blood-tissue samples, we study how systemic and local immune escape emerges. By integrating single-cell genomics, spatial transcriptomics, and patient-derived “avatar” mice, we reveal how immune evasion develops stepwise and identify windows where it remains optimally reversible. 

• Multimodal artificial intelligence for clinically actionable insights

  We use multimodal artificial intelligence to understand how the immune system links diverse clinical conditions. By integrating radiology, pathology, dermatology, gene expression, and immune profiling data with electronic health records, our models capture how systemic and tissue-level immune responses shape therapy outcomes and cancer risk. This framework goes beyond prediction, aiming to uncover immune mechanisms across diseases and enable precision prevention and treatment.