binomRegMethModelPred {SOMNiBUS}R Documentation

A smoothed-EM algorithm to estimate covariate effects and test regional association in Bisulfite Sequencing-derived methylation data

Description

This function returns the predicted methylation levels

Usage

binomRegMethModelPred(BEM.obj, newdata = NULL, type = "proportion")

Arguments

BEM.obj

an output from the function binomRegMethModel

newdata

the data set whose predictions are calculated; with columns 'Position', and covariate names that can be matched to the BEM.obj

type

return the predicted methylation proportion or the predicted response (in logit or other binom.link scale)

Value

This function returns the predicted methylation levels

Author(s)

Kaiqiong Zhao

Examples

#------------------------------------------------------------#
head(RAdat)
RAdat.f <- na.omit(RAdat[RAdat$Total_Counts != 0, ])
out <- binomRegMethModel(
  data=RAdat.f, n.k=rep(5, 3), p0=0.003307034, p1=0.9,
  epsilon=10^(-6), epsilon.lambda=10^(-3), maxStep=200, detail=FALSE,
  Quasi = FALSE, RanEff = FALSE
)
binomRegMethModelPred(out)

[Package SOMNiBUS version 1.1.0 Index]