BayesCpi is an implementation of Bayes Cpi to extend Bayes A and B for estimating direct SNP effects in
high dimensional data problems (p >> N). BayesCpi treats the prior probability, pi = P(SNP has zero effect), as unknown.
The C Function cBaysCpi is utilized for for speed ..
BayesCpi(ga, numiter = 5000, Pi = .9, y)A matrix of genotypes with a number of rows identical to the number of genotyped individuals and a number of columns identical to the number of SNPs. Values in the matrix are 0, 1, 2, & 5 for homozygous, heterozygous, other homozygous, & unknown genotypes, respectively.
Number of iterations
Proportion of SNP loci with 0 effect for Bayes C
Trait phenotypes or conventional breeding values
A list object with a vector of SNP estimates meanb and a vector of genomic values for individuals,
aHat are returned. It is also possible to extract the estimated number of SNP loci in nLoci.
This function runs Bayes C and Cpi to estimate direct SNP effects and the
proportion of loci with nonzero effects based on a matrix of genotypes,
ga and a vector of adjusted phenotypes, y, (Habier et al., 2011; BMC Bioinformatics 12:186).
As in other bayesian alphabet, Bayes Cpi is essential in high dimensional data problems with highly
overparameterized models (p >> N). It extends Bayes A and B to estimate the proportion of loci
with nonzero effect.
Other data management functions in gdmp can be used to construct the integer matrix of genotypes,
ga, to use as input to BayesCpi.
Habier et al. (2011). Extension of the bayesian alphabet for genomic selection. BMC Bioinformatics, 12, 186.