RM2: Regression models for localised mutations: Evaluating...

In reimandlab/RM2: Regression Models For Localised Mutations

Regression models for localised mutations: Evaluating differential mutation rates across classes of sites

Description

RM2() uses negative binomial regression to evaluate local mutational frequencies and processes between sites of the same class to flanking control regions

Usage

RM2(
  maf,
  sites,
  mut_class_columns = NA,
  cofactor_column = NA,
  window_size = 100,
  n_min_mut = 100,
  n_bin = 10
)

Arguments

maf	Data frame of mutations (prepared by get_mut_trinuc_strand) containing the following information: chr autosomal chromosomes as chr1 to chr22 and sex chromosomes as chrX and chrY start the start position of the mutation in base 1 coordinates end the end position of the mutation in base 1 coordinates ref the reference allele as a string containing the bases A, T, C or G alt the alternate allele as a string containing the bases A, T, C or G mut_trinuc trinucleotide context - where middle is C or T - with alternate allele mut_strand character indicating Watson (w) or Crick (c) ref_alt character indicating single-base substitution
sites	Data frame of site coordinates chr autosomal chromosomes as chr1 to chr22 and sex chromosomes as chrX and chrY start the start position of the mutation in base 0 coordinates end the end position of the mutation in base 0 coordinates
mut_class_columns	Character corresponding to the column(s) of mutation classes for grouped analysis
cofactor_column	Character corresponding to the column of binary cofactors
window_size	Integer indicating the half-width of sites and flanking regions (added to left and right for full width). (default 100)
n_min_mut	Integer indicating the minimum number of mutations required to perform analysis (default 100)
n_bin	Integer indicating the number of megabase bins to use (default 10)

Value

Data frame containing the regression estimates and likelihood ratio test output with the following columns: mut_type, pp, this_coef, obs_mut, exp_mut, exp_mut_lo, exp_mut_hi, fc, n_sites_tested

mut_type

A string identifying the mutation class

pp

The p-value from the likelihood ratio test

this_coef

The coefficient from is_site

obs_mut

The total number of observed mutations of that class

exp_mut

The expected number of mutations determined by the model

exp_mut_lo

Lower bound of 95% confidence interval

exp_mut_hi

Upper bound of 95% confidence interval

fc

Observed mutations divided by expected mutations

pp_cofac

The p-value from the likelihood ratio test of site:cofactor interaction

this_coef_cofac

The coefficient from the site:cofactor interaction term

n_sites_tested

The number of sites that were tested - all sites if no downsampling

reimandlab/RM2 documentation built on Aug. 13, 2022, 12:22 p.m.