Description

This track hub represents lncRNAs identified in Akerman et. al. (Cell Metabolism). 2611 human lncRNAs expressed in pancreatic beta cells were identified as described below. In brief, the transcripts were assembled by cufflinks based on 5 billion paired-end, directional reads from 29 islet samples. LncRNAs were retained if they had H3K4me3 enrichment at the 5’ end, were detected in FACS purified beta cells without evidence for enrichment in acinar cells, and were noncoding (CPAT score <0.364). We have also included previously annotated non-coding transcripts that were expressed in our RNA-seq data. Detailed procedure described below.

Display Conventions and Configuration

Track represents the exon structures of Human Islet lncRNAs (HI-LNCs). Human genome assembly: hg19.

Methods

De novo transcript assembly for islet lncRNAs

4,959 million stranded paired-end RNA-seq reads from 29 human islet samples were aligned to the human genome (hg19) as described (Moran et al., 2012). The reads falling into the top 143 most expressed genes were excluded. Remaining fragments were assembled into transcripts using cufflinks (Parameters: --min-isoform-fraction 0.01 --pre-mrna-fraction 0.10 --max-bundle-frags 5000000). Unannotated transcripts were extracted and retained if the following criteria were satisfied (i) presence of more than one exon in the transcript, (ii) presence of H3K4me3 enrichment (defined as described in Moran et al., 2012) in a region that is consistent with the transcriptional promoter (+1kb to -0.5 kb from the transcript 5 prime end), (iii) RNA Coding Potential Assessment Tool (CPAT) score < 0.364 (following recommendation for human genome sequences by (Wang et al., 2013)). For lncRNAs with multiple isoforms, the isoform with the longest exon number was kept, or the isoform with the longest exon length in the case of multiple isoforms with the same number of exons. Other annotated lncRNAs were added to the final compilation of non-coding transcripts if found not to overlap a novel non-coding transcript. These included Ensembl annotated lncRNA transcripts that were captured by Cufflinks, and lncRNAs defined by (Moran et al., 2012). Any lncRNA overlapping a coding exon on the same strand was removed from the list. Finally, the lncRNAs were filtered for expression of >0.05 RPKMs in FACS purified beta cells and an acinar to beta cell expression ratio < 3, using FACS purified beta and acinar RNA-Seq data processed as described in (Moran et al., 2012). This resulted in a total number of 2611 beta-cell enriched lncRNAs. Table S3 of Akerman et al represents the full list of lncRNAs identified in bed12 format.

Credits

Annotations were carried out by Dr. Nikolina Nakic, Delphine Rolando and Dr. Ildem Akerman, in J. Ferrer lab (Imperial College London and IDIBAPS)

Contact information

For inquiries, please e-mail Kim Cyrus

Further information

Jorge Ferrer Lab (Imperial College, UK) website

References

Akerman et al. 2016