INCOMMON:
Inference of Copy Number and Mutation Multiplicity in Oncology for Prognosis and Metastatic Tropism in 60,000 Clinical Cancer Samples
The intricate interplay between base substitutions and copy number alterations in cancer genomes critically influences tumor evolution and patient outcomes. Traditional studies often overlook this complexity by analyzing these biomarkers in isolation. In this presentation, I will introduce INCOMMON, a Bayesian framework developed to infer mutation multiplicity and copy number in over 60,000 pan-cancer samples using only targeted panel sequencing read counts and sample purity estimates. Leveraging MCMC sampling methods in Stan, the model estimates posterior probabilities for per-allele count rates, tumor purity, and mutation copy numbers, uncovering patterns of co-existing mutations and copy-number alterations. By stratifying over 20,000 patients based on composite genotypes across key oncogenes and tumor suppressor genes, we identify distinct prognostic groups with implications for survival and metastatic tropism that surpass conventional mutation-centric classifications. This integrated genomic approach deepens our understanding of cancer progression and metastasis, paving the way for novel biomarker discovery and clinical translation.