hwep (version 2.0.3)

rmlike: Likelihood inference for random mating

Description

Estimates gamete genotype frequencies using a maximum likelihood approach and runs a likelihood ratio test for random mating.

Usage

rmlike(nvec, thresh = 1, nstarts = 10)

Value

A list with the following elements:

p

The estimated gamete genotype frequencies. p[[i]] is the estimated frequency for gamete genotype i-1.

chisq_rm

The likelihood ratio test statistic for testing against the null of random mating.

df_rm

The degrees of freedom associated with chisq_rm.

p_rm

The p-value against the null of random mating.

Arguments

nvec

A vector containing the observed genotype counts, where nvec[[i]] is the number of individuals with genotype i-1. This should be of length ploidy+1.

thresh

All groups with counts less than nvec will be aggregated together.

nstarts

The number of random restarts to the EM algorithm. Set this to 0 for only one run.

Author

David Gerard

Details

Let q be the genotype frequencies. Let p be the gamete frequencies. Then random mating occurs if q == stats::convolve(p, rev(p), type = "open"). We test for this hypothesis using likelihood inference, while estimating p.

Examples

## Randomly generate gamete frequencies
set.seed(1)
ploidy <- 10
pvec <- stats::runif(ploidy / 2 + 1)
pvec <- pvec / sum(pvec)

## Genotype frequencies from gamete frequencies under random mating
qvec <- stats::convolve(pvec, rev(pvec), type = "open")

## Generate data
nvec <- c(stats::rmultinom(n = 1, size = 100, prob = qvec))

## Run rmlike()
rmlike(nvec = nvec)