Supplementary Figure 2: ARLNC1 is prioritized as a lineage-specific, cancer-associated lncRNA in prostate cancer.

(a) The top ten AR-regulated, localized prostate cancer-associated genes identified in Figure 2a, after applying an expression filter of at least four fold change (log₂FC = 2) upon DHT stimulation and at least 1 FPKM average expression in prostate cancer tissues. Genes are listed in the order of log₂-gene expression level difference between benign and localized prostate cancer tissues. (b) The top ten AR-regulated, metastatic prostate cancer-associated genes identified in Figure2b, after applying an expression filter of at least four fold change (log₂FC = 2) upon DHT stimulation and at least 1 FPKM average expression in prostate cancer tissues. Genes are listed in the order of log₂-gene expression level difference between benign and metastatic prostate cancer tissues. (c) Schematic illustration of the procedure used to nominate prostate lineage-specific, cancer-associated lncRNAs in prostate cancer. (d) The top 12 prostate tissue-specific, prostate cancer-associated lncRNAs identified in Figure 2c, after applying an expression filter of at least 10 FPKM in the prostate samples in the top 5% based on gene expression level. Genes are listed in the order of SSEA percentile (total n = 7,256 samples). (e) Relative expression (FPKM) of ARLNC1 across a panel of normal tissues in GTEx normal tissue RNA-seq cohort (n = 9,435 samples) (www.gtexportal.org). Box-plot definition: center, median; box limits, 1st and 3rd quartile; whiskers follow the 1.5 rule. (f) Tissue and cancer-specific expression of ARLNC1 according to MiTranscriptome. The SSEA test (total n = 7,256 samples) was used to determine whether each gene was significantly associated with a set of samples (e.g. prostate cancer) or cancer progression in a given lineage (e.g. prostate normal to prostate cancer). The genes were ranked according to their strength of association. Shown is the percentile rank for ARLNC1 (enrichment, positive; depletion, negative) across the tested sample sets/comparisons. (g) Oncomine concepts analysis of genes positively (top panel) or negatively (bottom panel) correlated with ARLNC1. The network was visualized using the Force-Directed Layout algorithm in the Cytoscape tool, with node sizes representing the number of genes in each molecular concept and node names indicating author of the primary study.