AMR_build: Build AMR vector

In jumentib/hdmax2: High Dimensionnal Mediation Analysis

Description

To build a vector for each AMR find with AMR_search

Usage

AMR_build(res, methylation, nb_cpg = 2)

Arguments

res	result object of AMR_search function
methylation	chromosomal position of markers (start)
nb_cpg	chromosomal position of markers (end)

Details

We use the series of pValues (one pValue per CpGs) obtained with the mEWAS regression method and the combination of pValue max2. To determine the potential AMRs used the combp method present in the ENmix package (Xu et al. 2016). This method uses the Fisher method to combine the pValues and also the base pair distance (bP) between CpGs (1000 bP maximum between nb_cpg CpGs on the same AMR). The information for each AMR is summarized by doing the mean (by row) of each CpG.

Value

A set of build AMRs.

AMR_mean contains the first components of each PCA for each DMR. CpG_for_each_AMR contains the list of markers (CpGs) present on each AMR.

Author(s)

Basile Jumentier

Examples

# library(hdma2)
#
# Run mEWAS
#
# res <- mEWAS(X = example$X, Y = example$Y, M = example$M, K = 5)
#
# Keep latent factors for mediation
#
# U <- res$U
#
# Run max2
#
# res <- max2(pval1 = res$pValue[, 1], pval2 = res$pValue[, 2])
#
# lauch AMR_search
#
# res <- AMR_search(chr = example$annotation$chr,
#                   start = example$annotation$start,
#                   end = example$annotation$end,
#                   pval = res$pval,
#                   cpg = example$annotation$cpg, nCores = 1)
#
# lauch AMR_build
#
# tmp <- AMR_build(res, methylation = example$M, nb_cpg = 2)

jumentib/hdmax2 documentation built on Feb. 25, 2022, 12:58 p.m.