Installation and Data Retrieval

Pleas follow the instructions provided in the protocol using MeDeCom.

Data import

Analogously to the general protocol, we use RnBeads for data import.

suppressPackageStartupMessages(library(RnBeads))
rnb.options(
  assembly = "hg19",
  identifiers.column = "submitter_id",
  import = T,
  import.default.data.type = "idat.dir",
  import.table.separator = "\t",
  import.sex.prediction = T,
  qc = T,
  preprocessing = F,
  exploratory = F,
  inference = F,
  differential = F,
  export.to.bed = F,
  export.to.trackhub = NULL,
  export.to.csv = F
)
sample.anno <- "annotation/sample_annotation.tsv"
idat.folder <- "idat/"
dir.report <- paste0("report",Sys.Date(),"/")
temp.dir <- tempdir()
options(fftempdir=temp.dir)
rnb.set <- rnb.run.analysis(dir.reports = dir.report,
                            sample.sheet = sample.anno,
                            data.dir = idat.folder)

Preprocessing and Filtering

The preprocessing and filtering is also performed in line with the general protocol using DecompPipeline.

suppressPackageStartupMessages(library(DecompPipeline))
data.prep <- prepare.data(rnb.set = rnb.set,
                          analysis.name = "TCGA_RefFree",
                          normalization = "bmiq",
                          filter.beads = TRUE,
                          min.n.beads = 3,
                          filter.intensity = TRUE,
                          min.int.quant = 0.001,
                          max.int.quant = 0.999,
                          filter.na = TRUE,
                          filter.context = TRUE,
                          filter.snp = TRUE,
                          filter.sex.chromosomes = TRUE,
                          filter.cross.reactive = TRUE,
                          execute.lump = TRUE)

Confounding factor adjustment

We use Independent Component Analysis (ICA) to adjust for the potential confounding factors age, race, gender, and ethnicity.

data.prep <- prepare.data(rnb.set = data.prep$rnb.set.filtered,
              analysis.name = "TCGA_RefFree",
              normalization = "none",
              filter.beads = FALSE,
              filter.intensity = F,
              filter.na = FALSE,
              filter.context = FALSE,
              filter.snp = FALSE,
              filter.sex.chromosomes = FALSE,
              filter.cross.reactive = FALSE,
              remove.ICA = TRUE,
              conf.fact.ICA = c("age_at_diagnosis","race","gender","ethnicity"),
              ica.setting = c("alpha.fact"=1e-5,"save.report"=TRUE,
                        "ntry"=10,"nmin"=20,"nmax"=50,"ncores"=10))

Feature selection

We select the 5,000 most variable CpGs across the samples for downstream analysis.

cg_subset <- prepare.CG.subsets(rnb.set=data.prep$rnb.set.filtered,
                                marker.selection = "var",
                                n.markers = 5000)

Deconvolution using RefFreeCellMix

We use the RefFreeCellMix function from the RefFreeEWAS package for deconvolution analysis. In contrast to MeDeCom, RefFreeCellMix does not use a regularization for the entries of the matrix to be at the extreme values 0 or 1. Thus, we just specify the number of components to be tested and the preprocessed object, along with the selected features.

md.res <- start.refreeewas.analysis(
  rnb.set=data.prep$rnb.set.filtered,
  cg.groups = cg_subset,
  Ks=2:15,
  factorviz.outputs=TRUE,
  work.dir = "TCGA_RefFree"
)

Downstream analysis

The resulting object is an object of type MeDeComSet, since the output from RefFreeCellMix has internally been transformed to the MeDeCom infrastructure. This enables loading the data set directly into the visualization tool FactorViz

suppressPackageStartupMessages(library(FactorViz))
startFactorViz(file.path(getwd(),"TCGA_RefFree","FactorViz_outputs"))