signature_exposure: Estimates the signature exposure of a mutational catalogue

In SigsPack: Mutational Signature Estimation for Single Samples

Description

Estimates the signature exposure of a mutational catalogue by reconstructing it from a chosen set of signatures, by default, the used method is quadratic programming

Usage

signature_exposure(mut_cat, P = get(utils::data("cosmicSigs", package =
  "SigsPack")), sig_set = NULL, FUN = decomposeQP, ...)

Arguments

mut_cat	matrix (f by n) conatining one or several mutational catalogues of samples whose signature exposures are to be estimated f is the amount of features that the profiles are defined on n is the number of catalogues
P	Matrix (f by k) containing the signature profiles of k signatures whose exposure is to be found k is the amount of signature profiles that P contains f is the amount of features that the profiles are defined on (default: COSMIC signature matrix 96 x 30)
sig_set	Numeric vector containing the index of the columns from the signature matrix (which are the signature profiles) that will be used reconstruct the mutational catalogue.
FUN	Function to estimate the signature exposure. Default: Quadratic programming (P has to be of full rank to use this method)
...	control parameters that will be passed to FUN

Value

List of the estimated signature exposure, the reconstructed profile of the sample, the cosine similarity between the two and the error

Examples

data(cosmicSigs)
signature_exposure(create_mut_catalogues(10,500)[['catalogues']])
signature_exposure(create_mut_catalogues(10,500)[['catalogues']], sig_set = c(2,7,16,28,30))
signature_exposure(as.matrix(create_mut_catalogues(10,500)[['catalogues']][,1]))

SigsPack documentation built on Nov. 8, 2020, 6:57 p.m.