SKAT {AssotesteR} | R Documentation |
SKAT is a regression method to test for association between genetic variants (common and rare) in a region. A score-based variance-component test.
SKAT(y, X, kernel = "linear", weights = NULL, a = 1, b = 25, perm = NULL)
y |
numeric vector with phenotype status: 0=controls, 1=cases. No missing data allowed |
X |
numeric matrix or data frame with genotype data coded as 0, 1, 2. |
kernel |
character string indicating the type of kernel to be used. Possible options are "linear", "wlinear", "quadratic", "IBS", "wIBS", "twowayx" ( |
weights |
optional numeric vector with weights for the genetic variants ( |
a |
positive numeric value for the parameter |
b |
positive numeric vallue for the parameter |
perm |
positive integer indicating the number of permutations ( |
The argument kernel
is used to specify the kernel function. "linear"
indicates the linear kernel, "wlinear"
indicates a weighted linear kernel, "quadratic"
indicates the quadratic polynomial kernel, "IBS"
indicates Identity-By-Share, "wIBS"
indicates weighted IBS, and "twowayx"
indicates a two-way interaction kernel.
For the weighted kernels ("wlinear"
and "wIBS"
), there are two options to get the weights. The default option (weights=NULL
) involves the calculation of the weights by taking into account the minor allele frequency of the variants. In this case, the weights are calculated from a Beta distribution with parameters a
and b
. The second option is to specify the weights by providing a vector of weights for the variants; in this case the length of the vector must equal the number of columns in X
. For more information see reference Wu et al (2011)
An object of class "assoctest"
, basically a list with the following elements:
skat.stat |
skat statistic |
asymp.pval |
asymptotic p-value of the applied statistic (distributed as chi-square with df=1) |
perm.pval |
permuted p-value |
args |
descriptive information with number of controls, cases, variants, permutations, and selected kernel |
name |
name of the statistic |
This method is computationally expensive
Gaston Sanchez
Wu MC, Kraft P, Epstein MP, Taylor DM, Chanock SJ, Hunter DJ, Lin X (2010) Powerful SNP-Set Analysis for Case-Control Genome-wide Association Studies. The American Journal of Human Genetics, 86: 929-942
Wu MC, Lee S, Cai T, Li Y, Boehnke M, Lin X (2011) Rare-Variant Association Testing for Sequencing Data with the Sequence Kernel Association Test. The American Journal of Human Genetics, 89: 82-93
## Not run: # load data genodata data(genodata) # phenotype (first column of genodata) pheno = genodata[,1] # genotype (rest of columns of genodata) geno = genodata[,-1] # apply SKAT with linear kernel myskat.linear = SKAT(pheno, geno, kernel="linear") myskat.linear # apply SKAT with weighted linear kernel # weights estimated from distribution beta(MAF, a=1, b=25) myskat.wlinear = SKAT(pheno, geno, kernel="wlinear", a=1, b=25) myskat.wlinear # apply SKAT with quadratic kernel myskat.quad = SKAT(pheno, geno, kernel="quadratic") myskat.quad # apply SKAT with IBS kernel myskat.ibs = SKAT(pheno, geno, kernel="IBS") myskat.ibs ## End(Not run)