DSS (version 2.12.0)

DMLtest.multiFactor: Perform statistical test for BS-seq data from general experimental design

Description

This function takes the linar model fitting results and performs Wald test at each CpG site, then return test statistics, p-values and FDR.

Usage

DMLtest.multiFactor(DMLfit, coef = 2)

Arguments

DMLfit

Result object returned from 'DMLfit.multiFactor' function.

coef

Integer to indicate which coefficient in the linear model is be tested for being zero. Be careful of intercept. If the model contains intercept, coef=2 indicate testing the first factor in the formula. If the model has no intercept, testing first factor should use coef=1.

Value

A data frame with following columns: chr, pos, stat, pvals, fdr. Each row is for a CpG site. Note that the CpG sites are sorted by chromosome and position.

Details

A Wald test is performed using the estimated coefficient and standard error from 'DMLfit.multiFactor'. P-values are derived from test statistics based on normal distribution. FDR is computed using cannonical Benjamini-Hochberg procedure.

See Also

DMLfit.multiFactor, DMLtest

Examples

data(RRBS)
## model fitting
DMLfit = DMLfit.multiFactor(RRBS, design, ~case+cell+case:cell)

## hypothesis testing
DMLtest.cell = DMLtest.multiFactor(DMLfit, coef=3)

## look at distributions of test statistics and p-values
par(mfrow=c(1,2))
hist(DMLtest.cell$stat, 100, main="test statistics")
hist(DMLtest.cell$pvals, 100, main="P values")