runPFA: Run PFA method to get the subtyping results.

In GaoLabXDU/CEPICS: CEPICS: a comparison and evaluation platform for integration methods in cancer subtyping

Description

Using PFA to apply dimension reduction and data integration, then using k-means to get the cluster results.

Usage

runPFA(data, maxk = 10, matlab)

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.
maxk	An integer value means the maximize number of clusters we will try.
matlab	MATLAB server variable when execute PFAinit() function.

Details

The author just provides the MATLAB code of the method. When we tried to transfer it to R, we found that the results of left division operation for non-square matrices in R and MATLAB were totally different. In order to compare and evaluate the original version of PFA, CEPICS uses R.matlab R package to initialize a MATLAB server to run the original MATLAB code. In addition, the original version code can only integrate three omics datasets. We improve the code and make it possible to integrate omics data without the number of datasets limit. You should execute PFAinit() to initialize a MATLAB server before running PFA, and close server after running.

References

Shi,Q. et al. (2017) Pattern fusion analysis by adaptive alignment of multiple heterogeneous omics data. Bioinformatics, 33, 2706-2714.

See Also

PFAinit

Examples

data(COAD_Methy)
data(COAD_miRNA)
data(COAD_mRNA)


matlab=PFAinit(path=getwd())
res=runPFA(data=list(COAD_Methy, COAD_miRNA, COAD_mRNA),maxk=10,matlab=matlab)
close(matlab)

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