Holon-S1 predicts genes and biomarkers from routine pathology slides, validated on unseen osteosarcoma case
iScienceLuvr · x · 2026-10-21
Holon introduced Holon-S1, a platform that predicts genes and biomarkers directly from routine pathology slides, showing its work by mapping genes and biological processes across the tumor instead of emitting a black-box score.
Key validation:
- In a retrospective N-of-1 case, the model — never trained on osteosarcoma — analyzed a patient's public slides and found fibroblast-associated genes (POSTN, COL3A1, FAP, etc.) overexpressed, matching what the patient's medical team had earlier confirmed with sophisticated molecular tests and targeted for treatment.
- It beat two other models at predicting spatial gene activity in a controlled comparison, with findings biologists can check rather than an opaque score.
More from Apps
- Fully code-generated video: Claude researches and scripts, NumPy music, MiniMax voice — technollama · 2026-10-21
- Stronger Model, Weaker Product: Dev Defaults to Fable 6.1 Sol Ultra Over 5.1 — curious_vii · 2026-10-21
- X creator subscriptions slammed: a price decrease stuck in approval for weeks — davidpattersonx · 2026-10-21
- Gemini APIs Add Blazing-Fast Live Speech Translation in Google AI Studio — DynamicWebPaige · 2026-10-21
- Ex-Google Brain/DeepMind team launches Vyb, an AI design model with editable layered output — testingcatalog · 2026-10-21
- Dev builds 'website as pop-up book' concept with Claude Opus 5.5 and Three.js — techartist_ · 2026-10-21