filter_genes: Filter genes considered for compute_localization by mutation...

In CamaraLab/TDA-TCGA: An R package for identification of relevant genetic alterations in cancer using topological data analysis

Description

Filter genes considered for compute_localization by mutation frequency, fraction of nonsynonymous mutations, and negative correlations between expression and mutation data

Usage

filter_genes(
  TDAmut_object,
  freq_threshold = 0.02,
  top_nonsyn_fraction = 350,
  upper_correlations_threshold = 0.9,
  lower_correlations_threshold = 1e-04
)

Arguments

TDAmut_object	object of class TDAmut with expression data, mutation data, nerve complexes
freq_threshold	threshold mutation frequency. Genes below this value are not considered in analysis. By default is 0.02
top_nonsyn_fraction	number of genes to keep with greatest nonsyn/nonsyn+syn fraction. By default is 350.
upper_correlations_threshold	upper q value theshold to keep genes displaying positive correlations between mutation and expression data. Genes with a median q value across all complexes exceeding this threshold are considered for further analysis. By default is 0.9.
lower_correlations_threshold	lower q value theshold to keep genes displaying negative correlations between mutation and expression data. Genes with a median q value across all complexes below this threshold are considered for further analysis. By default is 1e-4.

Value

Returns a TDAmut object populated with filtered genes to consider in compute_gene_localization. Optionally returns p and q values quantifying negative correlations between expression and mutation profiles of filtered genes.

CamaraLab/TDA-TCGA documentation built on Oct. 30, 2021, 5:14 p.m.