GBJ (version 0.5.4)

estimate_ss_cor: estimate_ss_cor.R

Description

Estimate the correlations between GWAS summary statistics using reference panel eigenvectors and reference panel genotypes.

Usage

estimate_ss_cor(ref_pcs, ref_genotypes, link_function)

Value

A list with the elements:

cor_mat

The d*d matrix giving the pairwise correlation of every two test statistics.

Arguments

ref_pcs

An n*m matrix containing PCs calculated from the reference panel. Here n is the number of subjects in the reference panel and m is roughly the number of PCs used in the original analysis which produced the summary statistics.

ref_genotypes

An n*d matrix holding the genotypes from the reference panel, where the d columns correspond to the d SNPs for which we have summary statistics. No missing data allowed.

link_function

Either "linear" or "logit" or "log".

Examples

ref_pcs <- matrix(data=runif(n=1000, min=-0.2, max=0.2), ncol=5)
ref_genotypes <- matrix(data=rbinom(n=2000, size=2, prob=0.3), ncol=10)
estimate_ss_cor(ref_pcs=ref_pcs, ref_genotypes=ref_genotypes, link_function="linear")