corrclusterMaker: corrclusterMaker

In xinchoubiology/Rcppsva: Rcpp Integration Surrogate Variable Analysis

Description

By clusterMaker, we can generate probe cluster under constraint of maximum distance. Adding the correlation constraint, we will truncate exsited cluster and calculate their correlation matrix sigma out

Usage

corrclusterMaker(dat.m = NULL, chr, pos, cluster = NULL, maxGap = 500,
  names, cutoff = 0.8, corrmat = NULL, cores = detectCores(),
  method = c("pearson", "spearman", "kendall"), merge = c("single",
  "complete", "average"))

Arguments

dat.m	n x m delta M|beta matrix for n CpG sites across 2*m paired different patient samples
chr	Chromosome vector
pos	position numeric vector
cluster	vector of probes clusters By distance(maxGap) constraints
maxGap	max gap length between 2 probes within a region ; Default = 500 also used in function clusterMaker
names	probe names vector ; used in function clusterMaker
cutoff	threshold to filter distance based cluster; Default = 0.8
corrmat	correlation matrix for each cluster; Default NULL
cores	number of thread used by corrclusterMaker
method	correlation calculation method; c("spearman", "pearson", "kendall"); Default "spearman"
merge	how to merge two sub-clusters; c("single", "complete", "average")
pos	CpG position with their probes id

Details

correlated cluster maker. Build the CpG cluster by correlation

only clusters contain >= 2 probes can get a corrected p-value

Value

list of clusterID vector

xinchoubiology/Rcppsva documentation built on May 4, 2019, 1:06 p.m.