cptModel: Computes within and between gene feature profile statistics...
In GISPA: GISPA: Method for Gene Integrated Set Profile Analysis

Description Usage Arguments Details Value Author(s) Examples

View source: R/cptModel.R

Computes the percentiles on the estimated profile statistics within a gene and across genes for one or more combination of feature or data types (expression, methylation, copy-number variation, or variant change)

1	cptModel(psm, genelist, cpt.data, cpt.method, cpt.max, profile, feature)

`psm`	: A data matrix of estimated gene profile statistics for each feature
`genelist`	: A vector of gene names or gene symbols corrosponding to the profile statistics
`cpt.data`	: Identify changepoints in the data using variance (cpt.var), mean (cpt.mean) or both (meanvar). Default is cpt.var.
`cpt.method`	: Choice of single or multiple changepoint model. Default is "BinSeg".
`cpt.max`	: The maximum number of changepoints to search for using "BinSeg" method. Default is 60. This number is dependent on the number of input data points
`profile`	: The desired direction of genomic change. The values are "up" (default) or "down" to select for increased or decreased gene set profile, respectively
`feature`	: Analysis type i.e., one ('1'), two ('2') or three ('3') dimensional feature analysis.

This function estimates within and between feature profile statistics by gens in addition to the summed percentiles and successive differences

Estimated change points in the input data set

Bhakti Dwivedi & Jeanne Kowalski

id <- 1000 ## number of probes
s <- 3 ## number of sample groups
dm <- matrix(runif(id*s,0,200), nrow=id, ncol=s, dimnames=list(paste("gene", 1:id, sep="") , paste("fs", 1:s, sep="")))
genelist <- rownames(dm)
cptModel(dm, genelist, cpt.data="var", cpt.method="BinSeg", cpt.max=60, profile="up", feature=1)