Molecular and transcriptome profiles based on CREBBP status. (A) Cytomolecular status and cooccurring mutational profile of patients based on CREBBP alterations, (B) Unsupervised clustering was performed across the entire transcriptome patient data set (n = 1079). Variance stabilizing transformation was performed (R function, vst), and scaled values were used to calculate the variance for each gene across all samples (R function, var). The genes were sorted based on variance, and the top 100 genes with highest variance were clustered (R function, ward.D2) and presented visually in a heatmap generated using pheatmap function in R.