outputToCGHregions: ADaCGH2 output as input to CGHregions

In ADaCGH2: Analysis of big data from aCGH experiments using parallel computing and ff objects

Description

Convert ADaCGH2 output to a data frame that can be used as input for CGHregions. This function takes as input the two possible types of input produced by the pSegment functions: either an ff object (and its associated directory) or the names of the RAM objects (the usual, in memory R objects) with the output, and chromosome, position, and probe name information.

Usage

outputToCGHregions(ffoutput = NULL, directory = getwd(),
                               output.dat = NULL,
                               chrom.dat = NULL,
                               pos.dat = NULL,
                               probenames.dat = NULL)

Arguments

ffoutput	The name of the ff object with the output from a call to a pSegment function. You must provide either this argument or all of the arguments output.dat, chrom.dat, pos.dat and probenames.dat.
directory	The directory where the initial data transformation and the analysis have been carried out. It is a lot better if you just work on a single directory for a set of files. Otherwise, unless you keep very carefull track of where you do what, you will run into trouble. This is only relevant if you use an ff object (i.e., if ffoutput is not NULL.)
output.dat	The name of the RAM object with the output from a call to a pSegment function. You must provide this argument (as well as chrom.dat, pos.dat and probenames.dat) OR the name of an ffobject to argument ffoutput.
chrom.dat	The name of the RAM object with the chromosome data information. See the help for inputToADaCGH.
pos.dat	The name of the RAM object with the position data information. See the help for inputToADaCGH.
probenames.dat	The name of the RAM object with the probe names. See the help for inputToADaCGH.

Value

A data frame of 4 + k columns that can be used as input to the CGHregions function. The first four columns are the probe name, the chromosome, the position and the position. The last k columns are the calls for the k samples.

Note

This function does NOT check if the calls are meaningfull. In particular, you probably do NOT want to use this function when pSegment has been called using merging = "none".

Moreover, we do not check if there are missing values, and CGHregions is likely to fail when there are NAs. Finally, we do not try to use ff objects, so using this function with very large objects will probably fail.

Author(s)

Ramon Diaz-Uriarte [e-mail]

See Also

pSegment

Examples

## Get location (and full filename) of example data file
## We will read from a text file

fnametxt <- list.files(path = system.file("data", package = "ADaCGH2"),
                         full.names = TRUE, pattern = "inputEx.txt")


##################################
#####
##### Using RAM objects
#####
##################################

## Read data into RAM objects



##    You might want to adapt mc.cores to your hardware
inputToADaCGH(ff.or.RAM = "RAM",
              textfilename = fnametxt,
              mc.cores = 2)

## Run segmentation (e.g., HaarSeg)

##    You might want to adapt mc.cores to your hardware
haar.RAM.fork <- pSegmentHaarSeg(cgh.dat, chrom.dat,
                                 merging = "MAD",
                                 mc.cores = 2)

forcghr <- outputToCGHregions(output.dat = haar.RAM.fork,
                              chrom.dat = chrom.dat,
                              pos.dat = pos.dat,
                              probenames.dat = probenames.dat)

## Run CGHregions
if(require(CGHregions)) {
  regions1 <- CGHregions(na.omit(forcghr))
  regions1
}

##################################
#####
##### Using ff objects
#####
##################################

if(.Platform$OS.type != "windows") {

## We do not want this to run in Windows the automated tests since
## issues with I/O. It should work, though, in interactive usage


## Create a temp dir for storing output.
## (Not needed, but cleaner).
dir.create("ADaCGH2_cghreg_example_tmp_dir")
originalDir <- getwd()
setwd("ADaCGH2_cghreg_example_tmp_dir")
## Sys.sleep(1)

##    You might want to adapt mc.cores to your hardware
inputToADaCGH(ff.or.RAM = "ff",
              textfilename = fnametxt,
              mc.cores = 2)


##    You might want to adapt mc.cores to your hardware
haar.ff.fork <- pSegmentHaarSeg("cghData.RData",
                                "chromData.RData",
                                merging = "MAD",
                                mc.cores = 2)

forcghr.ff <- outputToCGHregions(ffoutput = haar.ff.fork)

if(require(CGHregions)) {
  regions1 <- CGHregions(na.omit(forcghr.ff))
  regions1
}

### Clean up (DO NOT do this with objects you want to keep!!!)
load("chromData.RData")
load("posData.RData")
load("cghData.RData")

delete(cghData); rm(cghData)
delete(posData); rm(posData)
delete(chromData); rm(chromData)
unlink("chromData.RData")
unlink("posData.RData")
unlink("cghData.RData")
unlink("probeNames.RData")

lapply(haar.ff.fork, delete)
rm(haar.ff.fork)

### Delete all files and temp dir
setwd(originalDir)
## Sys.sleep(2)
unlink("ADaCGH2_cghreg_example_tmp_dir", recursive = TRUE)
## Sys.sleep(2)
}

ADaCGH2 documentation built on Nov. 8, 2020, 4:57 p.m.