AutoPipe (version 0.1.6)

TopPAM: Compute Top genes

Description

This function computes the n=TOP genes and the the best number of clusters

Usage

TopPAM(me, max_clusters=15,TOP=1000,B=100,clusterboot=FALSE)

Arguments

me

a matrix with genes in rows and samples in columns

max_clusters

max. number of clusters to check

TOP

the number of genes to take.

B

integer, number of Monte Carlo (<U+201C>bootstrap<U+201D>) samples.

clusterboot

A logical value indicating wether or not to calculate the Gap statistic and to bootstrap.

Value

a list of 1. A matrix with the top genes 2. A list of means of the Silhouette width for each number of clusters. 3. The optimal number of clusters. 4. gap_st the gap statistic of the clustering 5. best number of clusters according to the gap statistic.

Details

we use the clusGap algorithm from the package cluster to calculate the Gap statistic.

Examples

# NOT RUN {
##load the org.Hs.eg Library
library(org.Hs.eg.db)
#' ## load data
data(rna)
me_x=rna
res<-AutoPipe::TopPAM(me_x,max_clusters = 8, TOP=100,clusterboot=FALSE)
me_TOP=res[[1]]
number_of_k=res[[3]]

# }