Researchers at UCLA’s Jonsson Comprehensive Cancer Center developed an AI model that predicts patient outcomes in various cancer types using epigenetic factors (epifactors). This model categorizes cancers into distinct groups based on the gene expression patterns of these factors, which influence gene activation or deactivation. This method is more effective than traditional grading or epithelial-to-mesenchymal transition measures in predicting clinical outcomes.
The researchers analyzed 720 epigenetic factors in tumors from 24 different cancer types, finding significant outcome differences in 10 cancers. The AI model, trained on epigenetic factor gene expression levels, effectively divided patients into groups with differing survival chances.
This model, which overlaps significantly with cluster-defining signature genes, holds potential for broader applications and pediatric cancer analysis. Also, this study paves the way for therapies targeting epigenetic factors, such as histone acetyltransferases and SWI/SNF chromatin remodelers.
Ref: Cheng MW, Mitra M, Coller HA. Pan-cancer landscape of epigenetic factor expression predicts tumor outcome. Communications Biology. 2023;6(1):1-18. doi: https://doi.org/10.1038/s42003-023-05459-w
You must be logged in to post a comment.