Holon-S1 predicts genes and biomarkers from routine pathology slides across 16 organ types
iScienceLuvr · x · 2026-10-29
The iScienceLuvr team unveiled Holon-S1, an AI platform that predicts genes and biomarkers directly from routine pathology slides.
- Motivation: patients with the same diagnosis often have very different tumor biology, and molecular testing is slow and expensive
- Beat two other models in a controlled comparison on predicting spatial gene activity
- Identified potentially novel biomarkers in breast and colon cancer
- Evaluated across 16 organ types and 15+ scanners, on both FFPE and fresh-frozen slides
The team positions it as a way for researchers to mine biomarkers from slides they already have — a substantive AI-for-medicine advance.
More from Apps
- taan opens public beta: generate music from any multimodal input — ycombinator · 2026-10-29
- Free Grok Bot tool does the math on your health plan enrollment docs — alvelda · 2026-10-29
- Sakana AI launches enterprise Sakana API Platform in early access — SakanaAILabs · 2026-10-29
- Rian Johnson launches Chromagraph, an iPhone camera app that bypasses Apple's processing — petefang · 2026-10-29
- Apple's new touchscreen MacBook Pro costs up to $14,549 fully loaded — Polymarket · 2026-10-29
- Chrome's built-in AI powers on-device voice agents that talk to any web page — jason_mayes · 2026-10-29