Researchers led by UC San Diego and Trey Ideker unveiled MutationProjector, an AI-enabled tumor genome foundation model designed to interpret mutational landscapes and predict therapy responses by learning gene co-dependencies. The model was trained on genomic datasets spanning more than 30,000 tumors across 10 solid cancer types, using Project GENIE and The Cancer Genome Atlas. In reported tests, MutationProjector recovered masked mutation statuses and demonstrated performance against existing prediction methods for chemotherapy and immunotherapy response, including performance improvements based on interactions among multiple mutated genes rather than single-gene matches. Clinically, the approach targets a key bottleneck in precision oncology—limited rates of actionable matches from sequencing panels—and aims to translate multi-mutation context into more consistent therapeutic hypotheses.