Skip to content

README.md

Latest commit

 

History

History
98 lines (78 loc) · 3.68 KB

README.md

File metadata and controls

98 lines (78 loc) · 3.68 KB
Raw

Mutational Signature Exposures Across Ancestry Groups in Multiple Cancer Types

Overview

This project investigates whether mutational signature exposures differ by genomic ancestry across four cancer types (BRCA, LUAD, UCEC, PRAD) using TCGA data. We examine whether ancestry effects on mutational signatures are cancer-specific or generalizable across tumor types. This is a question with direct implications for health equity in precision oncology.

Research Question

Do patients of different genomic ancestries show significantly different mutational signature exposures within the same cancer type, and does this pattern generalize across cancer types?

Data Sources

Data files are not tracked in this repository. Download from:

  • UCSF + WashU Ancestry Calls: https://gdc.cancer.gov/about-data/publications/CCG-AIM-2020
  • MAF Files: GDC Data Portal, Cohort Level MAF for each cancer type
    • TCGA-BRCA (Breast) (rename to brca_maf.gz)
    • TCGA-LUAD (Lung) (rename to lung_maf.gz)
    • TCGA-UCEC (Uterine) (rename to uterus_maf.gz)
    • TCGA-PRAD (Prostate) (rename to prstate_maf.gz)

Place all files in the data/ folder before running scripts.

Project Structure

ancestry-print/
├── data/                          # not tracked in git
│   ├── UCSF_Ancestry_Calls.csv
│   ├── WashU_PCA_ethnicity_assigned.tsv
│   ├── brca_maf.gz
│   ├── lung_maf.gz
│   ├── uterus_maf.gz
│   └── prstate_maf.gz
├── scripts/
│   ├── merge_ancestry.R           # merge and validate ancestry calls
│   ├── extract_signatures.R       # mutational signature extraction
│   └── compare_groups.R           # cross-ancestry comparison and figures
├── results/
│   ├── kruskal_wallis_results.csv
│   ├── significant_ancestry_effects.csv
│   └── figures/
│       ├── fig1_boxplots_ancestry_signatures.png
│       ├── fig2_heatmap_ancestry_cancer.png
│       ├── fig3_cross_cancer_afr_eur_effect.png
│       └── fig4_mutation_burden.png
└── README.md

How to Run

Install required R packages:

install.packages("tidyverse")
install.packages("rstatix")
install.packages("ggpubr")

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("MutationalPatterns")
BiocManager::install("BSgenome.Hsapiens.UCSC.hg38")

Then run scripts in order:

source("scripts/merge_ancestry.R")
source("scripts/extract_signatures.R")
source("scripts/compare_groups.R")

Methods Summary

Genomic ancestry was inferred using published calls from Carrot-Zhang et al. (2020). For BRCA, a cross-validated approach was applied requiring concordance between UCSF and WashU ancestry calling methods (741 patients). For LUAD, UCEC and PRAD, UCSF calls were used as the primary source (681, 422 and 520 patients respectively). Cancer type was assigned based on MAF file membership. Somatic mutations were filtered to TCGA samples and single nucleotide polymorphisms only. Trinucleotide mutation matrices were constructed using MutationalPatterns and decomposed against COSMIC v3.3 reference signatures. Ancestry group differences in signature exposures were assessed using Kruskal-Wallis tests with Benjamini-Hochberg multiple testing correction.

Key Results

  • 2,364 patients across 4 cancer types with high-confidence ancestry assignments
  • SBS4 (tobacco) differences are LUAD-specific, suggesting cancer-specific ancestry effects
  • AFR prostate cancer patients show substantially higher mutation burden than EUR patients

Reference

Carrot-Zhang et al. (2020). Comprehensive Analysis of Genetic Ancestry and Its Molecular Correlates in Cancer. Cancer Cell, 37(5), 639-654.