runLRA: Run LRAcluster to get the subtyping results.

Description Usage Arguments Details Value References Examples

View source: R/runLRA.R

Description

We use LRAcluster R package to apply dimension reduction and data integration, then use k-means to cluster the samples and get subtyping results.

Usage

1
runLRA(data, type, maxdimension = 10, maxk = 10, cores = 1)

Arguments

data

A list of different omics data. Each data in data list should be format as a data matrix with rows representing features and columns representing samples.

type

A list of the types of different omics data. Each element of the list can be one of "binary","gaussian","poisson"

maxdimension

An integer value means the maximize dimension LRAcluster will tried from one, default 10.

maxk

the max number of clustering we will try.

cores

An integer value means the number of cores for parallel computing.

Details

ev is a metric that provided by the author of LRAcluster to determine the dimension number of best low dimension. We determine the best low dimensional data by the metric in LRAcluster. When using LRAcluster R package, the user must decide the best dimension number manually according to the explained variation plot in the dimension reduction step so that LRAcluster cannot automatically get the integrated results. Therefore, we propose a method to choose the number of dimensions automatically.

Value

Return a list with the following elements:

References

Wu,D. et al. (2015) Fast dimension reduction and integrative clustering of multi-omics data using low-rank approximation: application to cancer molecular classification. BMC Genomics, 16, 1022.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
data(COAD_Methy)
data(COAD_miRNA)
data(COAD_mRNA)


res=runLRA(data=list(COAD_Methy, COAD_miRNA, COAD_mRNA),
     type=list("gaussian","gaussian","gaussian"))

res=runLRA(data=list(COAD_Methy, COAD_miRNA, COAD_mRNA),
     type=list("gaussian","gaussian","gaussian"),maxdimension=10,maxk=10,iskmeans=TRUE)

GaoLabXDU/CEPICS documentation built on June 9, 2020, 2:31 a.m.