Foundation model and imaging link cancer genomes to scans

Researchers have developed a novel method that combines a foundation model called Evo~2 with clinical imaging to identify associations between genes and cancer phenotypes. This approach analyzes somatic mutations across three TCGA cohorts (clear cell renal cell carcinoma, hepatocellular carcinoma, and breast cancer) to predict a gene severity score without task-specific training. By correlating these scores with radiomic features from tumor segmentations, the method successfully identified established cancer drivers and uncovered 46 new genes, including those linked to Mendelian ciliopathy and cytoskeletal diseases, which were previously missed by conventional methods. AI

IMPACT This method could accelerate the discovery of new gene-phenotype associations in cancer research, potentially leading to new diagnostic or therapeutic strategies.

RANK_REASON This is a research paper detailing a new methodology for gene discovery in cancer. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• breast cancer
• hepatocellular carcinoma
• Mendelian ciliopathy
• renal clear cell carcinoma
• The Cancer Genome Atlas

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →