Skip to content

transcript.md

Latest commit

 

History

History
105 lines (74 loc) · 5.47 KB

transcript.md

File metadata and controls

105 lines (74 loc) · 5.47 KB
Raw

Transcript Models

← Back to eRNAkit

Overview

Enhancer RNAs (eRNAs) have traditionally been represented as genomic coordinates rather than defined RNA molecules. This limits the ability to investigate how individual eRNAs are processed, localised, structured or functionally regulated. The Transcript Models module provides a transcript-resolved annotation of human enhancer RNAs generated through pan-transcriptome reconstruction of stranded RNA-seq datasets across diverse human tissues, cell types and subcellular compartments. Using experimentally defined enhancer coordinates as a reference framework, RNA-seq reads were restricted to enhancer-associated regions and assembled to generate transcript models. The resulting catalogue resolves enhancer transcription into individual RNA molecules with defined boundaries, exon structures and strand information.

The resource provides:

  • Transcript-resolved eRNA GTF annotation.
  • Exon-exon junction BED files.
  • Cross-reference between transcript models and original eRNAkit enhancer coordinates.
  • Strand-resolved models for bidirectional enhancer transcription.
  • Annotation of transcript structures including multi-exonic eRNA isoforms.

These models enable RNA-centric analysis of eRNA biology, including regulation and function. We built this due to the Anene's Lab interest in post-trancriptional eRNA functions.

Standalone GTF Files

GTFs

File Description
eRNATmod_nchr.gtf Chromosomes are listed without the "chr" prefix, so 1, 2, 3 etc.
eRNATmod_withchr.gtf Chromosomes are listed with the "chr" prefix, so chr1, chr2, chr3 etc..

The GTF files is also include as a slot transcript inside the eRNAkitDB

GTF Annotation Information

The GTF follows standard GTF formatting, with transcript-specific information stored in column 9 attributes. Example: chr1 eRNAkit transcript 51001294 51005245 . + . gene_id “EN4528”; transcript_id “EN4528.1”; eRNAkit “en4528”;

Column 9 Attributes

This is a standard GTF column 9 formatting using key "value"; attribute pairs.

Attribute Description
gene_id Identifier assigned to the eRNA gene-level locus. Multiple transcript isoforms derived from the same reconstructed enhancer share the same gene identifier.
transcript_id Unique identifier assigned to each reconstructed transcript model.
eRNAkit Identifier linking reconstructed transcript models to the original eRNAkit enhancer annotation. Enables integration with eRNAkitDB expression, localisation and regulatory datasets.
gene_source Source annotation of the feature. eRNAkit indicates the enhancer locus annotation, while STRINGTIE indicates reconstructed transcript models.
gene_biotype Transcript classification. For eRNAkit annotations this is defined as eRNA.
cluster Identifier of the merged enhancer cluster where applicable. NA indicates no assigned cluster.
n_stranded Indicates whether the enhancer locus shows stranded transcription evidence and how many.
embedded_stranded Indicates whether transcript reconstruction supports strand-specific embedding within the enhancer region.
exon_number Position of each exon within the reconstructed transcript model.
cov Transcript coverage estimated during transcript reconstruction.
FPKM Fragments per kilobase of transcript per million mapped reads (FPKM) expression estimate.
TPM Transcripts per million (TPM) expression estimate.

Transcript Reconstruction Strategy

Transcript models were generated using:

  1. Curated enhancer coordinates from experimentally defined human enhancer resources.
  2. Stranded RNA-seq datasets representing diverse tissues, cell types and subcellular fractions (n=121).
  3. Read selection restricted to enhancer-associated genomic regions (enhancer cordinates are fixed).
  4. Pooled transcript assembly to maximise recovery of low-abundance eRNA molecules.
  5. Filtering to remove gene-associated transcription and unlikely transcript models.

The final catalogue provides:

  • 18,740 reconstructed eRNA loci.
  • 19,484 transcript models.
  • 1,538 reconstructed splice junctions.
  • Multi-exonic transcript structures representing processed eRNA species.

Biological Applications

The transcript models enable:

Application Description
Isoform-aware quantification Quantify individual eRNA transcript models rather than enhancer-level activity.
Splicing analysis Investigate exon connectivity and regulated splice-junction usage.
RNA localisation studies Analyse transcript-specific behaviour across cellular compartments.
RNA interaction studies Map RNA-binding proteins and RNA-RNA interactions to defined eRNA molecules.
Experimental design Design transcript-specific targeting strategies including primers, siRNAs and cloning constructs.

Integration with eRNAkitDB

Transcript models are integrated into eRNAkitDB through the eRNAkit identifier.

This enables direct connection between transcript-resolved models and:

  • eRNA expression profiles.
  • Subcellular localisation datasets.
  • RNA-RNA interaction maps.
  • RNA stability measurements.
  • Translation-associated datasets.

Citation

For the transcript-resolved enhancer RNA catalogue, please cite:
Benova, N., Kuklinkova, R., Ibenye, E., Boyne, J.R. and Anene, C.A., 2026. Reconstructing the human enhancer RNA transcriptome. bioRxiv, pp.2026-03.