modeling
Dynamic models & digital twins of disease
Dynamic “world” models of malignancy and digital twins of disease that trace how tumors evolve from normal tissue through invasive and metastatic disease across longitudinal patient cohorts.
Research
The Curtis Lab studies cancer as an evolving, dynamic ecosystem. We couple computational methods (AI/ML and agent-based models) with large-scale multi-modal profiling of patient cohorts and functional perturbations (genetic and pharmacologic) in living patient-derived tissue models, to forecast patient trajectories and build a causal understanding of disease. These approaches have produced new disease models, validated biomarkers, and therapeutic targets that we are advancing to the clinic.
multi-omics
AI/ML integration of multi-modal data
AI/ML methods to integrate and interpret multi-modal omics (genomics, spatial proteomics, transcriptomics and glycomics, and digital pathology) across real-world patient cohorts and clinical trials.
immunotherapy
Computational immunotherapy design
Novel immunotherapy strategies grounded in in-depth immune-phenotyping and computational modeling.
functional
Organoid & organ-on-a-chip perturbation
Patient profiles paired with organoid and organ-on-a-chip systems to functionally test which genetic and microenvironmental perturbations drive early malignant transformation.
platform
High-throughput perturbation & autonomous discovery
Robotics and platforms for high-throughput perturbation and phenotyping that feed our computational models, agentic workflows, and autonomous discovery loops — with the models, in turn, directing the next round of experiments.