LDparallel: Calculate Pairwise Linkage Disequilibrium with Multiple...

In merns/postgwas: GWAS Post-Processing Utilities

Description

Performs calculation of pairwise linkage disequilibrium for a number of SNPs on multiple CPU cores using the package genetics. This function is deprecated, used of the r2fast function of GenABEL is recommended (and much faster).

Usage

LDparallel(g, num.processes, ...)
LDparallel.genotype(g1, g2, ...)

Arguments

g	list. Has to contain genotype objects. Each list element represents a SNP with its associated genotypes.
num.processes	integer. The number of parallel processes to use. Defaults to the number of cores.
g1	object(genotype). Contains all genotypes for a SNP.
g2	object(genotype). Second SNP, like g1.
...	Arguments passed to the 'genetics' LD function

Details

This function offers thread-parallel calculation of linkage disequilibrium for a list of genotypes. The returned matrix of r-square correlation values is symmetric since LD(snp1, snp2) == LD(snp2, snp1). Uses a number of cores as determined by the multicore:::detectCores() and distributes (non-symmetric) SNP pairs evenly on them. The LDparallel.genotype is basically a copy of the original function LD.genotype from the package genetics but does not return P and chi^2.

Value

A matrix containing r-square values (numeric datatype). Row and column names match the SNP names provided in parameter g.

References

http://cran.r-project.org/web/packages/genetics/index.html

Examples

genotypes <- getGenotypes(
               snps = c("rs13240", "rs2479", "rs8411", "rs35681"), 
               gts.source = system.file("extdata", "example.gwaa", package = "postgwas"), 
               return.genotypes = TRUE 
             )
if(any(installed.packages()[, "Package"] == "parallel") && Sys.info()[["sysname"]] == "Linux") {
  LDparallel(genotypes)
} else {
  cat("Error: Mandatory package 'parallel' not installed or not supported by operating system.\n")
}

merns/postgwas documentation built on May 22, 2019, 6:53 p.m.