limma_screen: Screening procedure based on LIMMA

In nhejazi/methyvim: Targeted, Robust, and Model-free Differential Methylation Analysis

Description

Reduces the methytmle object by way of a limma model with empirical Bayes shrinkage to include only the CpG sites below the preset p-value cutoff. INTERNAL USE ONLY.

Usage

limma_screen(methytmle, var_int, type, cutoff = 0.05)

Arguments

methytmle	An object of class methytmle.
var_int	A numeric vector containing subject-level measurements of the variable of interest. The length of this vector must match the number of subjects exactly. If argument vim is set to "ate" or "rr", then the variable of interest is treated as an exposure, and the variable must be binary in such cases. If setting vim to target parameters assessing continuous treatment effects, then the variable need not be binary of course.
type	Character indicating the particular measure of DNA methylation to be used as the observed data in the estimation procedure, either Beta values or M-values. The data are accessed via minfi::getBeta or minfi::getM.
cutoff	Numeric indicating the p-value cutoff that defines which sites pass through the filter.

Value

An object of class methytmle, that is a subset of the original methytmle input object which has been reduced include only the relevant CpG sites. This subset is then used to estimate VIMs.

nhejazi/methyvim documentation built on April 30, 2020, 7:14 p.m.