AI-guided drug combination discovery for solid tumors
Most solid tumors evade single-agent therapies. EMphora predicts the target combinations that matter, trained on clinical trial outcomes across therapeutic modalities.
Get in Touch
Most solid tumors evade single-agent therapies. EMphora predicts the target combinations that matter, trained on clinical trial outcomes across therapeutic modalities.
Get in Touch
Tumors rapidly evolve around single-agent drugs. The combinations that prevent escape are not predictable from biology alone. They require learning from what has actually worked in patients across modalities.
EMphora identifies target pairs and combinations with synergistic clinical activity, validated on prospective held-out trial data. Modality-agnostic coverage across small molecules, antibodies, bispecifics, and T cell engagers.
Predicts synergistic target combinations for solid tumors by systematically mining and ranking evidence from published literature and clinical trial outcomes, stratified by validation level — from in vitro and in vivo to ex vivo and clinical. Combinations are classified by mechanism: pathway convergence, resistance circumvention, and synthetic lethality, rather than synergy scores that don't translate to patient outcomes. Modality-agnostic coverage across small molecules, antibodies, bispecifics, and T cell engagers.
Mi earned his PhD at Ruprecht-Karls-Universität Heidelberg under Prof. Julio Saez-Rodriguez, where his work on drug synergy prediction produced 10+ publications in journals including Nature Biotechnology, Cell Systems, and Genome Medicine. He organized the NCI-CPTAC DREAM Challenge, coordinating 100+ scientists to benchmark cancer proteomics prediction. At Stanford under Prof. Ash Alizadeh, he built computational frameworks for immuno-oncology drug discovery. Industry experience spans Sanofi and PrognomiQ, where he developed ML pipelines for oncology applications.
Interested in our platform or exploring partnership opportunities?