mutational-signatures

Analysis of Mutational Signatures

This analysis evaluates mutational signatures of the consensus SNV callers file.

The overall analysis considers two different approaches to assessing mutational signatures in OpenPedCan data:

• Evaluate single- and double-base substitution signatures from COSMIC signatures and Alexandrov et al, 2013 for all samples using deconstructSigs.

• Evaluate known (adult) CNS signatures, as identified in Degasperi et al, 2020 in samples using deconstructSigs

Usage

To run the mutational signature evaluations, use this in command line:

bash run_mutational_signatures.sh

This assumes you are in the top directory of the repository.

The analysis scripts will return results for each caller in the plots and results folder.

Contents

• The Rmd 01-known_signatures.Rmd runs single-base substitution (SBS) mutational signatures analyses using existing COSMIC and Alexander et al, 2013 signatures.

  • Calculates weights from deconstructSigs::whichSignatures which are are multiplied by each sample's sum of mutations as provided by deconstructSigs::mut.to.sigs.input.
  • These numbers are saved to the _signatures_results.tsv files in the results folder.

• The Rmd 02-cosmic_dbs_signatures.Rmd runs double-base substitution (DBS) mutational signature analyses using existing COSMIC v3.3.1 signatures, and calculates weights and saves results as is done in 01-known_signatures.Rmd.

• The script 03-fit_cns_signatures.R determines the relative contributions of 8 known (adult) CNS signatures in the OpenPedCan data (considers both WXS and WGS) using deconstructSigs::whichSignatures().

  • This script saves the fitted CNS signature exposures in results/fitted_exposures_signal-cns-deconstructSigs.RDS

• The Rmd 04-explore_hypermutators.Rmd investigates whether hyper-mutant and/or ultra-hypermutant samples have enrichment of signatures 3, 18, and/or MMR.

  • We also asked whether these samples show dysregulated TP53.

Summary of the calculations

Number of mutations per signature

The weights from deconstructSigs::whichSignatures are multiplied by each sample's sum of mutations as provided by deconstructSigs::mut.to.sigs.input. These numbers are saved to the _signatures_results.tsv files in the results folder.

For more information, see the calc_mut_per_sig code.

Proportion of tumors with a signature

A tumor is considered to have a particular signature if its weight is non-zero. This is divided by the number of tumor samples in that particular histology group. For more information, see the bubble_matrix_plot code.

Overall file structure

OpenPedCan-analysis
├── analyses
│   └── mutational-signatures
│       ├── 01-known_signatures.Rmd
│       ├── 01-known_signatures.nb.html
|       ├── 02-cosmic_dbs_signatures.Rmd
|       ├── 02-cosmic_dbs_signatures.nb.html
|       ├── 03-fit_cns_signatures.R
|       ├── 04-explore_hypermutators.Rmd
|       ├── 04-explore_hypermutators.nb.html
│       ├── README.md
│       ├── Rplots.pdf
|       ├── input
|       │   ├── COSMIC_v3.3.1_SBS_GRCh38.txt
|       │   ├── COSMIC_v3.3_DBS_GRCh38.txt
│       ├── plots
│       │   ├── cns
│       │   │   └── tmb_signature_dotplot.pdf
│       │   ├── cosmicv2
│       │   │   ├── bubble_matrix_cosmic_mutation_sig.png
│       │   │   └── signature_grouped_barplots
|       |   |   |   ├──barplot_{HISTOLOGY_GROUP}_cosmic_mutation_sig.png
|       |   |   |   └──...   
│       │   ├── cosmicv3
│       │   │   ├── bubble_matrix_cosmic_mutation_sig.png
│       │   │   └── signature_grouped_barplots
|       |   |   |   ├──barplot_{HISTOLOGY_GROUP}_cosmic_mutation_sig.png
|       |   |   |   └──...  
│       │   └── cosmicv3_dbs
│       │   |   ├── bubble_matrix_nature_mutation_sig.png
│       │   |   └── signature_grouped_barplots
|       |   |   |   ├──barplot_{HISTOLOGY_GROUP}_cosmicv3_dbs_mutation_sig.png
|       |   |   |   └──...   
│       │   └── nature
│       │   |   ├── bubble_matrix_nature_mutation_sig.png
│       │   |   └── signature_grouped_barplots
|       |   |   |   ├──barplot_{HISTOLOGY_GROUP}_nature_mutation_sig.png
|       |   |   |   └──...   
|       ├── results
|       │   ├── COSMICv2_signature_exposures.tsv
|       │   ├── COSMICv3.3_DBS_signature_exposures.tsv
|       │   ├── COSMICv3.3_signature_exposures.tsv
|       │   ├── Nature_signature_exposures.tsv
|       │   ├── cosmicv2_signatures_results.tsv
|       │   ├── cosmicv3.3_dbs_signatures_results.tsv
|       │   ├── cosmicv3.3_signature_results.tsv
|       │   ├── deconstructSigs_cns_exposures_merged.tsv
|       │   ├── fitted_exposures_signal-cns_deconstructSigs.rds
|       │   ├── hypermutator_sig_matrix.tsv
|       │   ├── nature_signatures_results.tsv
|       │   └── sig_matrix_by_molecular_subtype.tsv
│       ├── scripts
│       │   └── de_novo_signature_extraction.R
│       └── util
│           └── mut_sig_functions.R
├── data

Summary of custom functions

Function Name	Summary
sample_mut_sig_plot	Saves traditional mutational signature plots for each sample
calc_mut_per_sig	Given deconstructSigs::whichSignature output, formats the sample data into a data.frame and calculates the mutations per Mb for each sample and each signature
bubble_matrix_plot	Groups together data by histology and makes the bubble matrix plot
grouped_sig_barplot	Creates signature grouped barplots for histology group provided and only uses primary tumors' data