preds.at: Prediction using a possibly compacted tree

In Pigengene: Infers biological signatures from gene expression data

Description

A decision tree in Pigengene uses module eigengenes to build a classifier that distincuishes two or more classes. Each eigengene is a weighted average of the expression of all genes in the module, where the weight of each gene corresponds to its membership in the module. Each modules might contain dozens to hundreds of genes, and hence the final classifier might depend on the expression of a large number of genes. In practice, it can be desireable to reduce the number of necessary genes used by a decision tree. This function is helpful in observing changes to the classification output after removing genes with lower weights membership. It determines how a given decision tree would classify the expression data after removing a certain number of genes from consideration.

Usage

preds.at(c5Tree, pigengene, pos=0, Data)

Arguments

c5Tree	A decision tree that uses eigengenes from the pigengene object to classify the samples from the expression data.
pigengene	A object of pigengene-class, output of compute.pigengene
pos	Number of genes to be removed from the consideration. Genes are removed in ascending order of their absolute weight in the relevant modules. If 0 (default), the prediction will be done without compacting.
Data	The expression possibly new data used for classification

Value

A list with following components:

predictions	The vector of predictions after neglecting pos number of genes
eigengenes	The values for the eigenges after neglecting pos number of genes

See Also

Pigengene-package, pigengene-class, make.decision.tree, compact.tree, compute.pigengene, module.heatmap, get.used.features, get.fitted.leaf, Pigengene-package

Examples

     ## Data:
     data(aml)
     data(mds)
     data(pigengene)
     d1 <- rbind(aml,mds)

     ## Fiting the trees:
     trees <- make.decision.tree(pigengene=pigengene, Data=d1,
       saveDir="trees", minPerLeaf=15, doHeat=FALSE,verbose=3,
       toCompact=FALSE)
     preds1 <- preds.at(c5Tree=trees$c5Trees[["15"]], pigengene=pigengene, 
       pos=0, Data=d1)

Pigengene documentation built on Nov. 8, 2020, 6:47 p.m.