autoheatmap: Automated heatmap production for plotting cross-correlated...

In ttriche/groupedPMA: Grouped Penalized Multivariate Analysis -- additional penalties for PMA

Description

This function automates the production of heatmaps from supervised or unsupervised canonical correlation and principal component analyses.

Usage

autoheatmap(dat, design, x.covs, z.covs, y.cov=NULL, pen.x, pen.z, ...)

Details

autoheatmap is a convenience function for quickly dissecting rich datasets based on orthogonal decomposition of the cross-covariance matrices between various assays or covariates. Further details are provided in the vignette.

Examples

data(AML)

# methylation changes at homeobox gene promoters X mutation
HOXloci <- featureNames(AML)[which(fData(AML)$HOX==TRUE)]
CCA.HOX <- autoheatmap(AML, pen.x=0.5, pen.z=0.3, 
                       covs=colnames(AML.colorcoded), 
                       colorbar=AML.colorcoded,
                       loci=HOXloci)
AML.mutations[CCA.HOX$u<0] # DNMT3A, FLT3, NPM1

ttriche/groupedPMA documentation built on June 1, 2019, 2:51 a.m.