SignatureFit_withBootstrap_Analysis: Mutational Signatures Fit with Bootstrap Analysis

In Nik-Zainal-Group/signature.tools.lib: signature.tools.lib

Mutational Signatures Fit with Bootstrap Analysis

Description

This function is a wrapper for the function SignatureFit_withBootstrap_Analysis, which produces several plots for each sample in the catalogues cat. Fit a given set of mutational signatures into mutational catalogues to extimate the activty/exposure of each of the given signatures in the catalogues. Implementation of method similar to Huang 2017, Detecting presence of mutational signatures with confidence, which uses a bootstrap apporach to calculate the empirical probability of an exposure to be larger or equal to a given threshold (i.e. 5 This probability can be used to decide which exposures to remove from the initial fit, thus increasing the sparsity of the exposures. Note that SignatureFit_withBootstrap_Analysis will save the results of SignatureFit_withBootstrap in the outdir directory using the R save() function. If SignatureFit_withBootstrap_Analysis is rerun with the same setting, the saved file will be loaded to avoid rerunning the Signature Fit and figures will be replotted.

Usage

SignatureFit_withBootstrap_Analysis(
  outdir,
  cat,
  signature_data_matrix,
  nboot = 100,
  type_of_mutations = "subs",
  threshold_percent = 5,
  threshold_p.value = 0.05,
  method = "KLD",
  bf_method = "CosSim",
  alpha = -1,
  doRound = FALSE,
  nparallel = 1,
  n_sa_iter = 500
)

Arguments

outdir	output directory for the analysis, remember to add '/' at the end
cat	catalogue matrix, patients as columns, channels as rows
signature_data_matrix	signatures, signatures as columns, channels as rows
nboot	number of bootstraps to use, more bootstraps more accurate results
type_of_mutations	either "subs", "rearr" or "generic"
threshold_percent	threshold in percentage of total mutations in a sample, only exposures larger than threshold are considered
threshold_p.value	p-value to determine whether an exposure is above the threshold_percent. In other words, this is the empirical probability that the exposure is lower than the threshold
method	KLD or NNLS or SA
bf_method	bleeding filter method, one of KLD or CosSim, only if bleeding filter is used (alpha>-1)
alpha	set alpha to -1 to avoid Bleeding Filter
doRound	round the exposures to the closest integer
nparallel	to use parallel specify >1
n_sa_iter	set max Simulated Annealing iterations if method==SA
verbose	use FALSE to suppress messages

Value

returns the activities/exposures of the signatures in the given sample and other information, such as p-values and exposures of individual bootstrap runs.

Examples

res <- SignatureFit_withBootstrap_Analysis(catalogues,signature_data_matrix)

Nik-Zainal-Group/signature.tools.lib documentation built on April 13, 2025, 5:50 p.m.