MethylMix (version 1.6.0)

MethylMix: Mixture model for DNA methylation data in cancer.

Description

MethylMix identifies DNA methylation driven genes by modeling DNA methylation data in cancer vs. normal and looking for homogeneous subpopulations. In addition matched gene expression data (e.g. from microarray technology or RNA sequencing) is used to identify functional DNA methylation events by requiring a negative correlation between methylation and gene expression of a particular gene.

Usage

MethylMix(METcancer,METnormal,MAcancer,OutputRoot,Parallel)

Arguments

METcancer

This a matrix with the methylation data of cancer tissue with genes in rows and samples in columns

METnormal

This is a matrix with the normal methylation data of the same genes as in METcancer. Again genes in rows and samples in columns. The samples do not have to match with the cancer data.

MAcancer

This is the matched gene expression data for the same samples as in METcancer.

OutputRoot

Path to store the MethylMix results object.

Parallel

If true MethylMix will be run using parallel processing.

Value

MethylMixResults is a list with the following components:

MethylationStates

Matrix with for all genes the Methylation states using DM-value (i.e. Differential methylation values) that are defined as the methylation value with respect to the average normal methylation for a gene.

NrComponents

The number of methylation states for each gene.

Models

Beta mixture model parameters for each gene.

MethylationDrivers

Genes identified as functional and differential by MethylMix.

MixtureStates

A list with the DM-values for each gene that is functional and differential.

Examples

# load the three data sets needed for MethylMix
data(METcancer)
data(METnormal)
data(MAcancer)

# run methylmix on a small set of example data
MethylMixResults=MethylMix(METcancer,METnormal,MAcancer)

# try the parallel toolbox to speed up MethylMix modeling
MethylMixResults=MethylMix(METcancer,METnormal,MAcancer,Parallel=TRUE)