Leveraging Machine Learning to Map Neoantigen Hotspots and Guide Immunotherapy Across Cancers (Global Health Day 2025)

Le résumé fourni par la source

Neoantigens, arising from tumor-specific mutations, represent a promising aventue for precision immunotherapy by driving T-cell-mediated anti-tumor responses. Despite their potential, accurate identification of immunogenic neoantigens remains challenging due to high false-positive rates, tumor microenvironment complexity, and individual immune variability. To address these barriers, we enhanced ScanNeo2, a comprehensive workflow integrating genomic and transcriptomic data, with a machine learning–based predictor that improves neoantigen detection reliability and reduces false positives.We performed a large-scale, pan-cancer analysis across multiple cohorts spanning diverse cancer types and immunotherapy response profiles. Our study revealed shared neoantigen hotspots across tumor types, as well as population-specific mutation signatures that influence checkpoint blockade outcomes. This pan-cancer atlas reveals key determinants of immunotherapy response and supports strategies for patient stratification in diverse populations.Our findings underscore the global relevance of neoantigen-focused approaches to improving cancer immunotherapy outcomes.Presented at Global Health Day 2025 on 2025-11-19 (https://www.globalhealth.northwestern.edu/events/global-health-day/index.html)

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.