regress_against_covariate: Regress against covariates for trcQTL

In liangyy/mixqtl: Combine total and allele-specific counts for cis-QTL analysis

Description

Given total read count (trc, N x 1) and library size (lib_size, N x 1) along with covarites (N x C), regress covariates against log(trc / lib_size / 2) by two-step approach 1. Regress all covariates against response and select significant ones (alpha = 0.05) 2. Regress the selected covariates against response and return the predicted response

Usage

regress_against_covariate(trc, lib_size, covariates)

Arguments

trc	total read count (dimension = N x 1)
lib_size	library size (dimension = N x 1) If it is set to NULL, we consider trc as y and do not do any transformation.
covariates	covariates (dimension = C x N) with the first column indicating the covariate name

Value

predicted response based on the selected covariates (if no covariates are selected, fill in zeros)

Examples

regress_against_covariate(
  trc = rpois(100, 100),
  lib_size = rpois(100, 10000),
  covariates = cbind(
    data.frame(name = paste0('cov', 1:10)),
    as.data.frame(matrix(runif(1000), nrow = 10))
  )
)

liangyy/mixqtl documentation built on Sept. 17, 2020, 11:36 a.m.