Scientists Develop AI Reader For Cancer Hallmarks
Indian researchers from the S N Bose National Centre and Ashoka University published Dec 29, 2025 a study in Communications Biology introducing OncoMark, an AI framework that decodes cancer's molecular hallmarks. Trained on 3.1 million single cells across 14 cancer types and validated on nearly 20,000 patient samples, it achieved >99% internal and >96% external accuracy. The tool could help identify aggressive tumors and guide targeted therapies pending regulatory approval.
Key Points
- 1Analyzed 3.1 million single cells across 14 cancer types, creating pseudo-biopsies of hallmark activity
- 2Combined ten hallmarks to reduce noise, achieving >99% internal and >96% external accuracy
- 3Enables clinicians to identify aggressive tumors and target biological processes for personalised therapies
Scoring Rationale
Strong peer-reviewed validation and large-scale single-cell data support high impact, limited by required clinical regulatory approvals and deployment steps.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
