The abstract is read at the publisher; the summary is JClub's.
EAGLE-MUT predicts somatic single-base substitution probabilities in esophageal adenocarcinoma and melanoma that vary by several orders of magnitude with local DNA sequence context.
- Why it matters: Somatic mutations are the substrate for genome evolution toward age-related diseases, but their heterogeneous distribution is not well explained by sequence alone.
- What they did: The authors trained group DNA language models on sparse single base substitution data from esophageal adenocarcinoma (EAC) and melanoma (MELA), then fine-tuned per patient to compare predicted mutation distributions across genome elements.
- The result: In both cancers, T>G hotspots and coldspots align with nucleosome-related features and regulatory landmarks, and patient-specific fine-tuning yields distinct mutation patterns linked to genome function and potential driver mutations.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.