🔬 Training Transformers to solve 95% failure rate of Cancer Trials — Ron Alfa & Daniel Bear, Noetik

The failure of 90-95% of cancer drugs in clinical trials stems from poor patient selection rather than inadequate pharmacology. Noetik addresses this by building foundation models that analyze multimodal patient data—including H&E pathology stains, protein markers, and spatial transcriptomics—to identify therapeutically relevant cancer subtypes. By training these models on massive, proprietary datasets generated in-house, the team creates "world models" capable of simulating how specific patient populations respond to drug perturbations. This approach moves beyond simple, biased biomarkers, allowing for the discovery of novel targets and the optimization of clinical trial design. By leveraging self-supervised learning on high-density spatial data, these models provide a scalable, interpretable framework for matching the right molecule to the right patient, ultimately aiming to transform oncology from a trial-and-error process into a data-driven, predictive science.