Estimate pairwise inbreeding coefficient F by maximum likelihood

Estimates the inbreeding coefficient between all pairs of samples by maximum likelihood.

inbreeding_mle(
  x,
  f = seq(0, 1, l = 11),
  ignore_het = FALSE,
  report_progress = TRUE
)

Arguments

x

object of class mipanalyzer_biallelic.

f

values of f that are explored.

ignore_het

whether to ignore heterzygous comparisons, or alternatively call the major allele at every locus (see details).

report_progress

if TRUE then a progress bar is printed to the console.

Details

The probability of seeing the same or different alleles at a locus can be written in terms of the global allele frequency p and the inbreeding coefficient f, for example the probability of seeing the same REF allele is \((1-f)*p^2 + f*p\). This formula can be multiplied over all loci to arrive at the overall likelihood of each value of f, which can then be chosen by maximum likelihood. This function carries out this comparison between all pairwise samples, passed in as a matrix. The formula above only applies when comparing homozygous calls - for homo/het or het/het comparisons we can either ignore these loci (the default) or convert hets to homo by calling the major allele at every locus.