hardae/Mutational_Signatures

Project 7: Mutational Signatures in African Ancestry Cancer Patients # Mutational_Signatures # Project 7: Mutational Signatures in African Ancestry Cancer Patients **Author:** Karimat Adeola Busari **Portfolio:** Computational Cancer Genomics — African and Underserved Populations **Status:** Complete (v1.0) | Subgroup analysis in progress --- ## Biological Question What mutational processes are driving cancer in patients of African ancestry, and how do they compare to COSMIC reference signatures derived predominantly from Western cohorts? Mutational signatures are recurring patterns of DNA damage that reflect the underlying biological processes — environmental exposures, defective repair pathways, or endogenous cellular mechanisms — that have shaped a tumour's genome. By extracting these signatures using Non-negative Matrix Factorization (NMF) and matching them to the COSMIC v3.3.1 catalogue, we can identify which processes are most active in a given population. --- ## Why This Matters The COSMIC reference signatures were derived predominantly from European and Asian cancer cohorts. African and African-ancestry populations are severely underrepresented in global cancer genomic databases — meaning the mutational processes most relevant to African patients may be uncharacterised, misclassified, or simply missing from current reference catalogues. This project is designed as a reusable, population-aware pipeline that can be applied to any MAF-format cancer genomic dataset, with the explicit goal of eventually running on true continental African patient data as it becomes available through initiatives such as TACA (The African Cancer Atlas), H3Africa, and direct academic collaborations. --- ## Data Source **AACR Project GENIE v19.0** — accessed via Synapse (syn7222066) GENIE is a real-world cancer genomic registry from 19 leading cancer centres worldwide, containing somatic mutation data from over 200,000 patients across multiple cancer types. **Important population note:** GENIE's public cohort does not include African in …