expBATCH: Discovering and correcting batch effect.

In gnyamundanda/exploBATCH: Explore batch effect (exploBATCH) is a package for discovering and correcting for batch effect using an approach in Nyamundanda et al (2017).

Description

A tool for discovering and correcting for batch effect using an approach in Nyamundanda et al (2017).

Usage

expBATCH(D, batchCL, Conf = NA, mindim = 1, maxdim = 2, method = "ppcca", scale = "unit",SDselect=0)

Arguments

D	Expression dataset in samples by features format.
batchCL	A vector identifying batches.
Conf	A vector identifying biological classes of interest.
mindim	Minimum number of principal components to be considered. Default is 1.
maxdim	Maximum number of principal components to be considered. Default is 2. maxdim should be greater or equal to mindim.
method	Method for batch correction. Either PPCCA or ComBat. PPCCA is the default.
scale	Unit scale the data or none for no scaling. Unit scaling is the default
SDselect	Filter number of genes by setting SD value greater than zero using SDselect. SDselect=0 is the default.

Details

It searches for directions in the principal subspace that are associated with batch effect and correct for batch effect by subtracting this effect in the affected dimensions. Both the corrected and uncorrected dimensions are then used to predict the data.

Value

The output of this pipeline is saved in your working directory.

Author(s)

Nyamundanda, G., Poudel, P., Patil, Y. and Sadanandam, A.

References

Nyamundanda, G., Poudel, P., Patil, Y. and Sadanandam, A., 2017. A novel and robust statistical method to diagnose and correct batch effects in genomic data.

Examples

# Expression data and batch effect variable is required for this function. The first example in the Nyamundanda et al (2017) manuscript 
# requires the gene expression data for breast cancer samples, named Breast (which is in the exploBATCHbreast package on github), and 
# variable identifying batches named batchBreast (in the exploBATCH package) 
require(devtools)
install_github("gnyamundanda/exploBATCHbreast") # install_github requires devtools package
require(exploBATCHbreast)
data(Breast)
data(batchBreast)
expBATCH(D=Breast,batchCL=batchBreast,Conf=NA,mindim=2,maxdim=9,method="ppcca",SDselect=2)
# The second example in the manuscript requires the gene expression data for colon cancer samples, named Colon (which is in the exploBATCHcolon 
# package on github), and variables identifying batches and tumor status (biological effect), named batchColon and bioCL respectively, in the 
# exploBATCH package. 
#require(devtools)
#install_github("gnyamundanda/exploBATCHcolon") # install_github requires devtools package
#require(exploBATCHcolon)
#data(Colon)
#data(batchColon)
#data(bioCL) 
#expBATCH(D=Colon,batchCL=batchColon,Conf=bioCL,mindim=2,maxdim=9,method="ppcca",SDselect=2)

gnyamundanda/exploBATCH documentation built on May 20, 2019, 3:31 p.m.