Researchers published a multimodal AI framework intended to improve breast cancer prognostication, combining multiple data sources into a single predictive approach. The study, led by Witowski, Zeng, and Cappadona and published in Nature Communications, aims to integrate diverse inputs into a prognosis-oriented model designed for clinical applicability. A separate AI approach highlighted by HELIX focuses on RNA splicing prediction by integrating genomic sequence features with tissue-specific RNA binding protein (RBP) expression. While more mechanistic than purely prognostic, the splicing model provides a foundation for understanding disease-associated isoform shifts that can influence tumor behavior. Together, the updates reflect a push to move AI beyond single-modality imaging or genomics, toward multi-layer biological context for oncology decision-making and precision medicine research.