Eigenbeta: Eigenbeta
In xinchoubiology/Rcppsva: Rcpp Integration Surrogate Variable Analysis

Description Usage Arguments Details Value Author(s)

CpG sites are clustered by bump-hunter or combp. So the correlation(similarity) network will be represented by eigen unit(which is the 1st PC)

1
2
3

Eigenbeta(mset = NULL, cluster = NULL, nPC = 1, verbose = TRUE,
  softpower = 6, corr = c("pearson", "spearman", "kendall"),
  align = c("along average", ""))

`mset`	normalized methylation matrix of 485512 CpG sites
`cluster`	List; clustered neighboured CpG sites
`nPC`	number of principle components
`verbose`	Optional; TRUE(default), whether to print processing or not
`softpower`	6; For hub beta vector generation
`corr`	Correlation method; Optional; "pearson" default; "spearman" and "kendall"
`align`	Use the average beta(Methylation) value to control the orientation of eigenunit align = "along average"(default, means that eigenunit should be aligned with average beta value) align = "" (means letting the eigenunit vector as themselves)

get eigen-gene from pre-clustered prototype [Methods: BMC2007_Steve_Horvath] and the second-layer network analysis will be based on [eigengene network]

list eigenBeta eigen beta vectors' matrix for pre-clustered prototype units averageBeta average beta vectors for each clustered CpG unit

Xin Zhou xxz220@miami.edu

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