correlateTMBvalues: Correlate panel-based TMB values with WES-based TMB values

In acc-bioinfo/TMBleR: Tumor Mutational Burden Quantification From Gene Panels And WES

Description

This function takes in input TMB values quantified for the same samples from gene panels and from whole exome sequencing datasets and performs correlation analysis. Panel-based TMB values may come from real gene panel sequencing or from a simulated gene panel, generated using the simulatePanel function.

Usage

correlateTMBvalues(panel.TMB, WES.TMB, corr.coeff, title.plot)

Arguments

panel.TMB	a numeric vector of length n containing TMB values estimated from a gene panel for n samples
WES.TMB	a numeric vector of length n containing TMB values estimated from whole exome sequencing for n samples
corr.coeff	type of correlation coefficient: "pearson", "kendall" or "spearman"
title.plot	a string describing plot title

Value

Returns a pdf file with correlation plot, the Spearman's correlation coefficient with 95 number of samples.

Author(s)

Laura Fancello

Examples

# Read vcf files containing somatic mutations identified by WES -------------
data(ExampleWESvcfs)

# Read files with WES and gene panel design ---------------------------------
data(ExampleWESdesign)
data(ExamplePaneldesign)

# Filter for gene panel simulation ------------------------------------------
# Remove known cancer mutations
vcfs_NoCancer_ForPanel <- applyFilters(vcfs = ExampleWESvcfs, 
                                       assembly = "hg19", 
                                       design = ExamplePaneldesign, 
                                       remove.cancer = TRUE, 
                                       tsList = NULL, 
                                       variantType = NULL)

# Filter for original WES ---------------------------------------------------
# Filter out synonymous mutations
vcfs_NoSynonymous_WES <- applyFilters(vcfs = ExampleWESvcfs, 
                                      assembly = "hg19", 
                                      design = ExampleWESdesign, 
                                      remove.cancer = FALSE, 
                                      tsList = NULL, 
                                      variantType = c("synonymous"))

# Subset the WES dataset so that it will only contain variants in the regions
# targeted by the panel you want to simulate
SimulatedPanel_NoCancer <- applySimulatePanel(WES = vcfs_NoCancer_ForPanel, 
                                              WES.design = ExampleWESdesign,
                                              panel.design = ExamplePaneldesign, 
                                              assembly = "hg19")

# TMB quantification --------------------------------------------------------
# Perform TMB quantification on the simulated panel
TMBs_SimulatedPanel <- applyTMB(inputForTMB = SimulatedPanel_NoCancer, assembly = "hg19")
# Perform TMB quantification on the original Whole Exome sequencing
TMBs_WES <- applyTMB(inputForTMB = vcfs_NoSynonymous_WES, assembly = "hg19")

# Correlate WES-based and simulated panel-based TMB values ------------------
cor_plot <- correlateTMBvalues(panel.TMB = TMBs_SimulatedPanel$Tot_Number_Mutations, 
                                   WES.TMB = TMBs_WES$Tot_Number_Mutations, 
                                   corr.coeff = "spearman",
                                   title.plot="Correlation panel-based and WES-based TMB")
# Visualize plot of Spearman correlations
print(cor_plot)

acc-bioinfo/TMBleR documentation built on Dec. 18, 2021, 10:21 p.m.