We use the package ‘pheatmap’ to represent a hierarchical clustering of the variables selected by the multilevel analysis (for method = ‘splsda’) as it can display the factor(s) cond.
Usage in mixOmics
data(data.simu); X.simu <- data.simu$X stimulation <- data.simu$stimu repeat.simu <- data.simu$sample result.1level <- multilevel(X.simu, cond = stimulation, sample = repeat.simu, ncomp=3, keepX=c(200,200,200), tab.prob.gene=NULL, method = 'splsda') pheatmap.multilevel(result.1level, col_sample=col.sample, col_stimulation=col.stimu, label_annotation=NULL,border=FALSE,clustering_method="ward", show_colnames = FALSE,show_rownames = TRUE,fontsize_row=2)
See also Multilevel:Liver Toxicity case study for a two factors display.