blupga_CAND: BLUP|GA using SNPs in Candidate Genes

In dkainer/BLUPGA: BLUP|GA genomic prediction model

Description

This function runs the BLUP|GA method where SNPs in candidate genes are weighted in the GRM prior to GBLUP

Usage

blupga_CAND(G, phenodata, valset, GRMs = NULL, verbose = T)

Arguments

G	G-matrix constructed using all available SNPs and samples Defaults to NULL.
valset	vector of indices that defines which rows in phenodata will be set to NA and used for cross-validation Defaults to NULL.
GRMs	List of G-matrices, each constructed from just the SNPs in one candidate gene/region. Defaults to NULL.
@param	phenodata data frame with 2 or 3 columns. One col must be named 'ID' and contain sample IDs. Another col must be named 'y' and contain the phenotypes. If fixed effects are included then a 3rd col called 'FE' should contain the categorical effects. Defaults to NULL.

Value

A data frame containing the correlation between the predicted phenotype and the true phenotype of the individuals in the valset.

Since BLUP|GA is run for each value of omega (W) from 0.0 to 1.0 in increments of 0.10, each row of the returned data frame contains the cross-validation correlation at one value of omega (W). This allows the user to find the value of W at which the predictive ability (COR) is maximised.

W

omega weighting for selected SNPS in candidate genes (0.0–1.0)

COR

cross validation predictive ability (0.0–1.0)

Examples

blupga_CAND()

dkainer/BLUPGA documentation built on Jan. 3, 2020, 1:09 a.m.