| grammar {GenABEL} | R Documentation |
Fast approximate test for association between a trait and genetic polymorphisms, in samples with genetic sub-structure (e.g. relatives). The function implements several varieties of GRAMMAR ('gamma','gc', and 'raw').
grammar(polyObject, data,
method = c("gamma", "gc", "raw"), propPs = 1, ...)
polyObject |
object returned by
|
data |
object of |
method |
to be used, one of 'gamma','gc', or 'raw' |
propPs |
proportion of non-corrected P-values used
to estimate the inflation factor Lambda, passed directly
to the |
... |
arguments passed to the function used for
computations, ( |
With 'raw' argument, the original GRAMMAR (Aulchenko et al., 2007) is implemented. This method is conservative and generates biased estimates of regression coefficients.
With 'gc' argument, the GRAMMAR-GC (Amin et al., 2007) is implemented. This method solves the conservativity of the test, but the Genomic Control (GC) lambda is by definition "1" and can not serve as an indicator of goodness of the model; also, the estimates of regression coefficients are biased (the same as in 'raw' GRAMMAR).
GRAMMAR-Gamma (default 'gamma' argument) solves these problems, producing a correct distribution of the test statistic, interpretable value of GC Lambda, and unbiased estimates of the regression coefficients. All together, the default 'gamma' method is recommended for use.
Object of scan.gwaa-class
Gulnara Svischeva, Yurii Aulchenko
GRAMMAR-Raw: Aulchenko YS, de Koning DJ, Haley C. Genomewide rapid association using mixed model and regression: a fast and simple method for genomewide pedigree-based quantitative trait loci association analysis. Genetics. 2007 Sep;177(1):577-85.
GRAMMAR-GC: Amin N, van Duijn CM, Aulchenko YS. A genomic background based method for association analysis in related individuals. PLoS One. 2007 Dec 5;2(12):e1274.
GRAMMAR-Gamma: Svischeva G, Axenovich TI, Belonogova NM, van Duijn CM, Aulchenko YS. Rapid variance components-based method for whole-genome association analysis. Nature Genetics. 2012 44:1166-1170. doi:10.1038/ng.2410
# Using clean ge03d2 data require(GenABEL.data) data(ge03d2.clean) # take only a small piece for speed ge03d2.clean <- ge03d2.clean[1:200,] # estimate genomic kinship gkin <- ibs(ge03d2.clean[,sample(autosomal(ge03d2.clean),1000)], w="freq") # perform polygenic analysis h2ht <- polygenic(height ~ sex + age, kin=gkin, ge03d2.clean) h2ht$est # compute mmscore stats mm <- mmscore(h2ht, data=ge03d2.clean) # compute grammar-gc grGc <- grammar(h2ht, data=ge03d2.clean, method="gc") # compute grammar-gamma grGamma <- grammar(h2ht, data=ge03d2.clean, method="gamma") # compare lambdas lambda(mm) estlambda(mm[,"chi2.1df"]) lambda(grGamma) estlambda(grGamma[,"chi2.1df"]) lambda(grGc) estlambda(grGc[,"chi2.1df"]) # compare top results summary(mm) summary(grGamma) summary(grGc)