strand_bias_test: Significance test for strand asymmetry

In MutationalPatterns: Comprehensive genome-wide analysis of mutational processes

Description

This function performs a two sided Poisson test for the ratio between mutations on each strand. Multiple testing correction is also performed.

Usage

strand_bias_test(strand_occurrences, p_cutoffs = 0.05, fdr_cutoffs = 0.1)

Arguments

strand_occurrences	Dataframe with mutation count per strand, result from 'strand_occurrences()'
p_cutoffs	Significance cutoff for the p value. Default: 0.05
fdr_cutoffs	Significance cutoff for the fdr. Default: 0.1

Value

Dataframe with poisson test P value for the ratio between the two strands per group per base substitution type.

See Also

mut_matrix_stranded, strand_occurrences, plot_strand_bias

Examples

## See the 'mut_matrix_stranded()' example for how we obtained the
## following mutation matrix.
mut_mat_s <- readRDS(system.file("states/mut_mat_s_data.rds",
  package = "MutationalPatterns"
))

tissue <- c(
  "colon", "colon", "colon",
  "intestine", "intestine", "intestine",
  "liver", "liver", "liver"
)

## Perform the strand bias test.
strand_counts <- strand_occurrences(mut_mat_s, by = tissue)
strand_bias <- strand_bias_test(strand_counts)

## Use different significance cutoffs for the pvalue and fdr
strand_bias_strict <- strand_bias_test(strand_counts,
  p_cutoffs = 0.01, fdr_cutoffs = 0.05
)

## Use multiple (max 3) significance cutoffs.
## This will vary the number of significance stars.
strand_bias_multistars <- strand_bias_test(strand_counts,
  p_cutoffs = c(0.05, 0.01, 0.005),
  fdr_cutoffs = c(0.1, 0.05, 0.01)
)

MutationalPatterns documentation built on Nov. 14, 2020, 2:03 a.m.