GWAF (version 2.2)

glmm.lgst: function for testing association between a dichotomous trait and a genotyped SNP in family data using Generalized Linear Mixed Effects model

Description

Fit a Generalized Linear Mixed Effects model (GLMM) with logistic link and a normal distributed random intercept for each cluster to test association between a dichotomous phenotype and one genotyped SNP in a genotype file with user specified genetic model. Each family is treated as a cluster. When analyze rare variants for dichotomous traits, GLMM, as implemeted by this function, is recommended over other methods such as GEE. The trait-SNP association test is carried out by the lmer function from package lme4. This function is called in glmm.lgst.batch function to apply association test to all SNPs in the genotype data.

Usage

glmm.lgst(snp, phen, test.dat, covar = NULL, model = "a")

Arguments

snp

genotype data of a SNP

phen

a character string for a phenotype name in test.dat

test.dat

the product of merging phenotype, genotype and pedigree data, should be ordered by "famid"

covar

a character vector for covariates in test.dat

model

a single character of 'a','d','g', or 'r', with 'a'=additive, 'd'=dominant, 'g'=general and 'r'=recessive models

Value

Please see output in glmm.lgst.batch.

Details

The glmm.lgst function tests association between a dichtomous trait and a SNP from a dataset that contains phenotype, genotype and pedigree data (test.dat), where the dataset needs to be ordered by famid.

References

Bates D, Maechler M and Bolker B. (2010) lme4: Linear mixed-effects models using S4 classes Longitudinal data analysis using generalized linear models. http://cran.r-project.org/web/packages/lme4/.

Chen MH, Liu X, Wei F, Larson MG, Fox CS, Vasan RS and Yang Q. (2011) A comparison of strategies for analyzing dichotomous outcomes in genome-wide association studies with general pedigrees. Genetic Epidemiology 35:650-657.

See Also

lmer function from package lme4

Examples

## Not run: 
# glmm.lgst(snp=data[,"rs123"],phen="CVD",test.dat=data,model="a",covar=c("age","sex"))
# ## End(Not run)