 The MOMA platform is a machine learning approach that can accurately predict the survival rate of colorectal cancer patients based on their histological patterns, multi-omics data, and clinical profiles. The platform also identifies interpretable pathology patterns that are correlated with gene expression profiles, microsatellite instability status, and clinically actionable genetic alterations. This information can be used to guide treatments for colorectal cancer patients. This article was offered by Pei Chien Tsai, Tsung Hua Li, Kuang Kai Kui, and others.