Figure 1: Proliferative and invasive cellular states in melanoma biopsies and cultures.
From: Decoding the regulatory landscape of melanoma reveals TEADS as regulators of the invasive cell state
(a) Non-negative matrix factorization on TCGA-SKCM RNA-seq data results in three sample clusters. (b) Functional characteristics of two states revealed by GSEA on the invasive and the proliferative meta-rankings integrated across SKCM RNA-seq and two microarray compendia, using various sources of functional data (L) Literature; (R) Reactome; (KEGG) KEGG pathways; (GO) Gene Ontology; (T) TargetScan. (c) Expression heatmap for TCGA samples showing a core subset of invasive and proliferative gene signatures (the GSEA overlap between the Hoek signatures and our ranking). The samples are ranked according to MITF expression, and the expression levels of both MITF and ZEB1 are indicated on top of the heatmap. Below the heatmap are mosaic plots of several samples. Each mosaic shows the expression of all variable genes in a 25 × 26 grid, whereby each field contains one or more genes. Genes and clusters with similar expression profiles across the cohort are placed close to each other in the grid. The mosaics show a high similarity among the invasive samples, and a strong difference between invasive and proliferative samples. SKCM is RNA-seq data from TCGA; Hoek et al.16, is microarray data from melanoma cultures; Compendium A and B are melanoma microarray data from GEO (see Supplementary Table 1 for accession numbers used).
