NMF_Extraction: NMF_Extraction

In FunGeST/Palimpsest: Package

Description

Extracts mutational signatures de novo using NMF. Also estimates the optimal number of mutational signatures in the input.

Usage

NMF_Extraction(
  input_matrices = NULL,
  range_of_sigs = 1:20,
  num_of_sigs = "auto",
  nrun = 10,
  method = "brunet",
  plot_sigs = TRUE,
  resdir = NA
)

Arguments

input_matrices	Palimpsest input list of mutation number and proportion matrices.
range_of_sigs	Numerical range of signatures. If "num_of_sigs" is set to auto, NMF will estiamte the optimal number of signatures in the input within this range before extracting with this number.
num_of_sigs	The number of mutational signatures to extract. If left to the default "auto" value, the appropriate number of signatures will be estimated from NMF metrics (which are plotted in resdir). "auto" is recommended on the first run, but once the optimal number of signatures is known, setting this argument to that value will make the extraction quicker.
method	Specification of the NMF algorithm. ‘brunet’ method corresponds to the standard NMF algorithm from Brunet et al. (2004, PNAS).
plot_sigs	Logical, if TRUE (default), the extracted mutational signatures will be plotted in the resdir.
resdir	Results directory.

Examples

SBS_denovo_signatures <- NMF_Extraction(input_matrices = SBS_input, range_of_sigs = 1:20,nrun = 10,resdir = resdir)

FunGeST/Palimpsest documentation built on June 10, 2021, 6:35 p.m.