Skip to main content
Fig. 3 | Genome Biology

Fig. 3

From: Intergenic disease-associated regions are abundant in novel transcripts

Fig. 3

Functional properties of captured transcripts. a Comparison of tissue-specific expression of captured transcripts to lncRNAs, pseudogenes, and protein-coding genes (Illumina Body Map), as measured by Tau tissue specificity index (0 for broadly expressed, 1 for tissue specific genes) [79]. b Heatmap of tissue-specific captured transcripts (τ > 0.8) across tissues. Unsupervised clustering performed on τ components (1-Expression/max(Expression)), colored by tissue specificity from low (white) to high (red). Statistically significantly non-randomly clustered branches after 10,000 bootstraps, as calculated by Pvclust [115] are marked by red rectangles. **p value of a cluster branch < 10–3. c Enrichment of genomic regions of captured transcripts for known FANTOM enhancers. Log odds ratios (ORs) of enrichment (with 95% confidence intervals) compared to lncRNAs, pseudogenes, and protein-coding genes. Genomic regions of both introns and exons were included in the analysis. FANTOM enhancers in red, randomized regions in blue. d Enrichment of GWAS SNPs in transcript regions. Log OR of enrichment for GWAS SNPs (p value < 5 × 10-8), compared to intronic regions. Exons in red, promoters in yellow, 3’ UTRs in blue. Hollow circles denote enrichment for common SNPs. Statistically significant adjusted p values (Χ2 test, p values < 0.05) are denoted with asterisks. e Example of a captured transcript with independently validated function. Transcript GCS1669 overlaps three known lncRNAs, with CCAT1 being functionally validated in liver and prostate carcinogenesis. Gray box marks captured region. Previously observed splice sites are denoted in red. f Expression levels of transcript GCS1669 across tissues

Back to article page