dksCustomClass: Create a classification object from predefined gene...

In dualKS: Dual KS Discriminant Analysis and Classification

Description

This utility function will build a DKSClassifier object from your own list of gene ids for use by dksClassify. This is useful if you want to use the classification funtionality of this package, but already have gene signatures you want to use (as opposed to generating them with dksTrain.

Usage

dksCustomClass(upgenes=NULL, upclass=NULL, downgenes=NULL, downclass=NULL)

Arguments

upgenes	A vector of gene ids that are upregulated in the classes of intereset. Note that upgenes and downgenes need not both be specified, but at least one must be.
upclass	A vector or factor of classes specifying which signature each gene in upgenes belongs to. (Note that upgenes can describe an arbitrary number of classes, and the order of upgenes is not important so long as the position of ids in upgenes and the position of classes in upclass correspond to one another).
downgenes	A vector of gene ids that are downregulated in the classes of intereset. Note that upgenes and downgenes need not both be specified, but at least one must be.
downclass	A vector or factor of classes specifying which signature each gene in downgenes belongs to. (Note that downgenes can describe an arbitrary number of classes, and the order of downgenes is not important so long as the position of ids in downgenes and the position of classes in downclass correspond to one another).

Value

An object of class DKSClassifier.

Author(s)

Eric J. Kort, Yarong Yang

See Also

dksTrain, dksSelectGenes, dksClassify, DKSGeneScores, DKSPredicted, DKSClassifier

Examples

	data("dks")
	up <- rownames(exprs(eset))[1:300]
	cls <- factor(rep(c(1,2,3), 100))
	classifier <- dksCustomClass(upgenes=up, upclass=cls)

dualKS documentation built on Nov. 8, 2020, 8:30 p.m.