ONC-001: Cancer Immunotherapy

The Resistance Problem

Immune checkpoint inhibitors (ICIs) targeting PD-1/PD-L1 and CTLA-4 have transformed oncology, generating over $30B annually. Yet response rates remain disappointingly low: only 20-40% in most solid tumors, with many responders eventually developing resistance. The fundamental challenge is that tumors employ multiple, redundant immune evasion mechanisms that monotherapies cannot overcome.

AI-Discovered Immune Escape Pathways

Our biological superintelligence analyzed spatial transcriptomics, single-cell sequencing, and proteomics data across multiple cancer types. Rather than searching for known immune checkpoints, we trained our model to identify entirely novel mechanisms by which tumors suppress T cell infiltration and activation.

The AI discovered five previously uncharacterized immune evasion pathways, each independently capable of conferring ICI resistance. Our digital patient models predict that non-responding tumors frequently activate multiple pathways simultaneously—explaining why single-agent ICIs fail.

Rational Combination Design

Using digital patient models, we simulated 10¹² potential drug combinations targeting these five pathways. Our AI predicted synergistic combinations that not only block multiple escape routes but create positive feedback loops—where inhibiting pathway A enhances the effects of targeting pathway B.

Three lead small molecule combinations emerged with promising predicted response profiles. Each combination has been validated in our digital patient platform using hundreds of tumor microenvironment models, demonstrating consistent predicted efficacy across genetic backgrounds and tumor subtypes.