Pan-cancer proteogenomics signatures associated with HRD, MSI, APOBEC, and smoking.

This is a repo with the scripts for CPTAC pan-cancer proteogenomics signatures project. Also see here for other CPTAC pan-cancer papers:

• https://doi.org/10.1016/j.cell.2023.07.014

• https://doi.org/10.1016/j.cell.2023.07.013

• https://doi.org/10.1016/j.ccell.2023.06.009

Analysis scripts:

1. Define_High_Low_groups -- this directory contains scripts that are used to define "High" and "Low" groups of samples for each mutational signature based on that signature mutational count and fraction.

2. Expression_markers -- this directory contains scripts to run differential expression marker analysis between "High" and "Low" samples groups using cancer cohort as a covariate.

   • The analysis was done using pan-cancer harmonised dataset, that was also used in the 1st and 2nd CPTAC papers listed above.

   • In the folder, there are three scripts for marker calculation based on: gene, protein and phosphosite expression.

3. Proteogenomic_markers_Figs_5_6 -- this directory contains analysis and plotting scripts for analysis of proteogenomic markers associated wih tht HRD, MSI, APOBEC, and smoking cancer phenotypes.

   • Proteogenomic_markers_Figs_5_6/Volcano_plots -- scripts for plotting protein and phosphosite expression markers. Top markers are highlighted in those plots.

   • Proteogenomic_markers_Figs_5_6/Heatmap_plots -- scripts for plotting top gene, protein and phosphosite expression, that are also used for subtyping using ConsensusClusterPlus R package. For each of the phenotype/cancer pair, there are two steps/scripts. For example, for HRD-BRCA pair the scripts should be used in the following order:

     • Proteogenomic_markers_Figs_5_6/Heatmap_plots/Generate_matrix.BRCA.20240105.R -- script to select top markers, perform subtyping based on those, and also creating the matrix and annotations needed for plotting.

     • Proteogenomic_markers_Figs_5_6/Heatmap_plots/Make_heatmap_BRCA_HRD.20240105.R -- script to make a heatmap plot with ComplexHeatmap R package, using the results from the previous step.