Transcriptomic landscape of esophageal squamous cell carcinoma (ESCC). (A) Unsupervised hierarchical clustering analysis of genome-wide RNA transcripts in ESCC and adjacent normal tissue using a microarray with 7,419 long noncoding RNAs (lncRNAs) and 27,958 coding RNAs. A self-organizing map (SOM) of either whole transcriptome (both lncRNAs and mRNAs) or lncRNAs or mRNA was produced from each group samples using gene expression dynamic inspector (GEDI). Mosaic patterns are pseudo-colored SOMs to show integrated biological entity in each sample. The color gradient from red to blue color indicates the expression level from high to low. (B) Liner models for microarray data analysis (limma) identified differentially expressed lncRNAs (DE-lncRNAs) and coding genes (DE-coding gene) in ESCC relative to normal tissue. FDR < =0.01, P <0.001. DE-lncRNAs and DE-coding gene expressions are displayed in genome view (hg19) with RCircos plot. In each circle view, blue line: downregulation, red line: upregulation.