GenABEL (version 1.8-0)

cocohet: Test for compound heterozygote effects

Description

Detecting rare recessive and compound heterozygote alleles in genome wide association.

Usage

cocohet(data, trait, window, return_all_result = TRUE, makePlot = FALSE, test = "CHI2", min_expected_cut_off = -1)

Arguments

data

Genotype data for analysis. Object of class snp.data

trait

Vector with binary trait data. Object of class integer or numeric.

window

Number of SNPs on the "right" of a given SNP which are used in analysis with a SNP. Object of class integer

return_all_result

If FALSE then return only a vector where each element is a chisq obtained as a maximum chisq between a given SNP and SNPs on the right within a window. If TRUE then return also a matrix where chisq's for all tests are stored. Object of class logical

makePlot

whether the Manhattan-type plot should be produced (TRUE or FALSE)

test

Name of the test to be performed. Available tests are "CHI2", "YATES" (chi2 with Yates correction), and "FISHER". Object of class character

min_expected_cut_off

In case this is >=0 and test is NOT Pearson's chisq test then Pearson's chisq test (!) is performed only for SNPs which produce acontingency table where the expected number of subjects in each field is >min_expected_cut_off. Otherwise the specified test is performed. Object of class integer or numeric

Value

A list is returned.

Details

The function is an inplementation of the method aimed to detect a gene-phenotype association caused by recessive and compound heterozygote genotype states of multiple rare variants at a particular gene locus. This method is described in 'Detecting Low Frequent Loss-of-Function Alleles in Genome Wide Association Studies with Red Hair Color as Example'; Fan Liu, Maksim V. Struchalin, Kate van Duijn, Albert Hofman, Andre G. Uitterlinden, Yurii S. Aulchenko, and Manfred Kayser. PLoS ONE 6(11): e28145. doi:10.1371/journal.pone.0028145

The three tests are implemented: Pearson's chi-square test, Pearson's chi-square test with Yates correction, Fisher exact test. In case when the input parameter min_expected_cut_off is <0 the="" choosen="" in="" input="" parameter="" "test"="" test="" is="" performed.="" if="" min_expected_cut_off="">= 0 then always Pearson's chi-square test is performed exept of the cases when expected number of subjects in a field of contingency table is

References

Fan Liu, Maksim V. Struchalin, Kate van Duijn, Albert Hofman, Andre G. Uitterlinden, Yurii S. Aulchenko, and Manfred Kayser. Detecting Low Frequent Loss-of-Function Alleles in Genome Wide Association Studies with Red Hair Color as Example'. PLoS ONE 6(11): e28145. doi:10.1371/journal.pone.0028145

Examples

require(GenABEL.data)
data(srdta)
chis2_nocorrection <- cocohet(data=gtdata(srdta),
trait=phdata(srdta)$bt, window=3, test="CHI2")