A multimodal framework for comprehensive driver variant prediction in cancer

Le résumé fourni par la source

Cancer genomes contain many mutations, but only a subset drive tumor development. Accurately pinpointing these driver variants remains challenging. We aim to build an accurate and interpretable model by combining DNA sequence, protein 3D structure, and cancer omics data. We present ModVAR, a multimodal model that integrates DNA sequences, predicted protein tertiary structures, and cancer omics data to classify driver variants. The approach uses pre-trained models (DNAbert2 and ESMFold) and a self-supervised strategy for cancer omics profiles. We evaluate performance on clinically and experimentally validated driver variants with standard classification metrics, examine therapeutic relevance through molecular docking, assess modeling of variants in intrinsically disordered protein regions, and analyze modality contributions. Here we show that ModVAR achieves strong accuracy across benchmarks for identifying validated driver variants. It prioritizes variants with potential therapeutic actionability supported by docking analyses, and the inclusion of structural predictions enables effective modeling of variants in intrinsically disordered regions. Interpretation indicates that the protein structure modality contributes most to predictions. At scale, the method produces 3,971,946 publicly available variant annotations. ModVAR integrates sequence, structure, and cancer omics signals to aid driver-variant discovery. It provides robust performance across tasks, supports hypothesis generation and target discovery, and supplies a large-scale resource that advances cancer research and personalized therapy. Many genetic changes are found in cancers, but only a few drive the disease. Our aim is to spot these important changes more accurately. We built ModVAR, a computer tool that combines three types of information: the DNA sequences, the structures of protein, and cancer omics data. We tested ModVAR on variants (DNA changes) checked in clinics and experiments. It identifies likely driver variants and highlights changes that could be suitable for treatment, supported by simulations of how drugs might bind to proteins. We share an open resource with about 3.97 million annotated variants to help research and, in time, support more precise patient care. Yang, Chen et al. present ModVAR, a multimodal tool that integrates DNA sequence, protein structure, and cancer omics to predict cancer driver variants. It performs strongly on validated benchmarks, flags potentially actionable variants supported by docking, and models changes in intrinsically disordered protein regions.

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