University of California San Diego researchers introduced MutationProjector, an AI-enabled tumor genome foundation model designed to map mutational co-dependencies and predict therapy response. The approach compresses genomic variance into an interpretable representation to support actionable feature discovery for targeted treatments and immunotherapy resistance. The model was trained on more than 30,000 tumors across 10 solid cancer types using resources including Project GENIE and The Cancer Genome Atlas. In evaluations, it recovered masked gene statuses and matched or exceeded existing prediction methods across multiple cohorts. The update targets a practical bottleneck in precision oncology: interpreting multi-mutation panels where most patients lack direct matching to approved therapies due to limited actionable coverage.