locus_overlap: Probability of random null overlap among groups for outlier...

In j-a-thia/genomalicious: A smorgasbord of R functions for population genomic analyses

Probability of random null overlap among groups for outlier loci

Description

This functions calculates the null probability of observing overlap in outlier loci between 2+ groups with respect to a scenario whereby loci are drawn at random from a set of loci.

Usage

locus_overlap(lociList, lociSig, perms = 1000)

Arguments

lociList	List: A list of loci analysed for each group. Each index is a group, containing a vector loci names.
lociSig	List: A list of significant loci for each group. Each index is a group, containing a vector of loci names.
perms	Integer: The number of permutations to run for significance testing. Default = 1000

Details

This non-parametric style test is not a strict population genetic statistical method per se, but provides an informal way to compare parameters of an outlier analysis (number of groups, number of analysed loci, and the number of observed outlier loci) to random chance.

It is very important that the order of indices in lociList and lociSig are in the same, because each index is expected to correpond to a group.

The hypothesis being tested is that the number of observed loci overlapping among groups is greater than that observed due to random chance.

It is assumed that the amount of overlap is across all groups. E.g. if only 2 groups are provided, then a locus must be significant in both groups. If 3 groups are provided, then a locus must be significant in all 3 groups.

Permutations are run where loci are randomly drawn from each group's set of loci at a size equivalent to that deemed statistically significant in empircal tests. In other words, if a test determined 2 loci out of 100 were significant, a null draw would randomly sample 2 loci from the set of 100.

Value

The value returned is the proportion of null permutations that were greater than the observed value (i.e. the empirical number of significant loci shared between groups).

Examples

library(genomalicious)

grps <- list(A=c(1:100), B=c(20:100), C=c(20:100))

sig <- list(A=c(8, 10, 16, 67, 68, 69, 88, 89, 90), B=c(22, 50, 51, 56, 57, 88, 95, 96), C=c(20, 21, 23, 44, 60, 70, 75, 88, 100))

locus_overlap(lociList=grps, lociSig=sig, perms=1000)

j-a-thia/genomalicious documentation built on Oct. 19, 2024, 7:51 p.m.