run_mcmc: Run the MCMC for full sibling inference

In eriqande/fullsniplings: Full sibling reconstruction from SNP data

Description

Run the MCMC for full sibling inference

Usage

run_mcmc(genos, mu = 0.005, pair_prob_cutoff = 0.001, burn_in, num_sweeps)

Arguments

burn_in	Number of sweeps to discard as burn-in. (In one sweep, every individual in the sample has been updated once by Gibbs sampling.)
num_sweeps	Number of sweeps after burn in to use as a sample for MCMC.
genos	The data frame of SNP genotypes
mu	Genotyping error rates per locus (recycles as necessary)
pair_prob_cutoff	the fraction of simulated full sib pairs that should have a log-likelihood ratio for the sibling relationship less than the cutoff that will be used. (The value pair_prob_cutoff is used to determine the which pairs have a high enough log-likelihood ratio to be considered as possible full siblings.)

Examples

# this is a much shorter run than one should do:
run_mcmc(fs_dev_test_data$chinook_full_sibs_genos, burn_in = 3, num_sweeps = 5)

eriqande/fullsniplings documentation built on May 16, 2019, 8:45 a.m.