Code - Exonic enhancers are a widespread class of dual-function regulatory elements

This is the Github repository of ExonEnhancer. 

Full Changelog: https://github.com/benoitballester/ExonEnhancer/commits/v2

This repository contains the scripts used for data curation, analyses and figures generation [Zenodo Data] in our manuscript: "Exonic enhancers are a widespread class of dual-function regulatory elements." 

In this study, we explore the role of exonic regions in gene regulation in four species. We identify and characterize many protein-coding exons as candidate exonic enhancers (cEEs), a previously underappreciated class of cis-regulatory elements embedded within exons. By integrating TF ChIP-seq, chromatin accessibility data (DNase-seq/ATAC-seq), high-throughput enhancer-reporter assays (STARR-seq, luciferase), and CRISPR-based validations, we show that exonic enhancers (EEs) play crucial roles in gene regulatory networks while retaining their protein-coding function.

Supplementary Data is available on Zenodo.

1. Identification of EEs Across Multiple Species

   • Systematic discovery of EEs using TF ChIP-seq, chromatin accessibility, and STARR-seq data.
   • Many protein-coding exons exhibit enhancer activity.

2. Dual Coding and Regulatory Roles

   • EEs retain protein-coding functions while simultaneously acting as cis-regulatory elements.
   • Both synonymous and nonsynonymous variants can disrupt EE activity and downstream gene expression.

3. Long-Range Interactions and Target Gene Regulation

   • Promoter capture Hi-C and eQTL analyses confirm interactions between EEs and gene promoters.
   • CRISPR-based inactivation of EEs demonstrates regulatory effects on both host and distal target genes.

4. Clinical and Evolutionary Implications

   • Pan-cancer (TCGA) analyses show that mutations in EEs correlate with altered gene expression and clinical outcomes.
   • Evolutionary conservation indicates that EEs are functionally constrained yet contribute to species-specific regulatory innovations.

Below is a brief description of each folder, organized by thematic for easy navigation and reproducibility :

• EE_selection/
  Scripts used to define candidate exonic enhancers (cEEs) based on TF ChIP-seq peaks and additional filtering criteria.

• Control_selection/
  Procedures for generating negative/positive control sets, ensuring unbiased comparisons with cEEs.

• Chromatin_accessibility/
  Scritps to curate DNase-seq, ATAC-seq and histones marks datasets assessing open chromatin in exons and cEEs across multiple species.

• Conservation_and_structure/
  Scripts for multi-species and pairwise alignment, phyloP scores, AlphaFold predictions, MobiDB disorder rates and gene-age analyses for cEEs.

• Randomisation/
  Scripts used to validate enrichment within cEEs through randomization and permutation tests.

• TFBS_in_EE/
  Motif analysis pipeline (e.g., JASPAR-based TFBS predictions) overlapping with candidate exonic enhancers.

• STARR-seq_experiment/
  Analysis scripts for STARR-seq data, including the reads pipeline and SNPs analysis.

• STARR-seq_catalog/
  Data curation of STARR-seq peaks from public data sources, as well as cEEs biotype signature definition.

• G-quadruplex/
  Scripts examining G4-forming sequences (G-quadruplex) in cEEs vs. control exons.

• Interaction_data/
  Integration of promoter capture Hi-C, eQTL (GTEx), and ENCODE-rE2G resources to identify robust cEE–target gene interactions.

• gnomADv3_analysis/
  Variant filtering/annotation pipelines and constraint analyses for common variants in gnomAD v3 that intersect with candidate exonic enhancers.

• GWAS_analysis/
  Overlaps of known GWAS loci with cEEs to reveal potential trait- and disease-associated variants within coding enhancer regions.

• PanCancer_analysis/
  Curation and analysis scripts (differential expression and survival analysis) for the TCGA PanCanAtlas.

• Genes_specificity/
  Scripts used to infer the cEEs host/target genes expression tendencies.

• CRISPRi_plots/
  Scripts used to generate the CRISPRi figures.

• UCSC_trackhub/
  Configurations for easily visualizing cEEs, TF binding, and variant positions in the UCSC Genome Browser.

1. Environment Requirements
   • Scripts were primarily run on Python (>= 3.9) and R (>= 4.0).
   • Required dependencies include common genomic libraries (e.g., Bioconductor packages in R like GenomicRanges) and Python packages like pandas, numpy, and pybedtools.

If you use this dataset, software, or any derived resources, please cite:

Mouren, J.-C., Torres, M., van Ouwerkerk, A., Manosalva, I., Gallardo, F., Spicuglia, S., & Ballester, B.
“Supplementary Data — Exonic enhancers are a widespread class of dual-function regulatory elements.”
Zenodo. https://doi.org/10.5281/zenodo.15079251

And the corresponding manuscript:

Mouren, J.-C., Torres, M., van Ouwerkerk, A., Manosalva, I., Gallardo, F., Spicuglia, S., & Ballester, B.
“Exonic enhancers are a widespread class of dual-function regulatory elements.” (Manuscript in preparation.)

For questions regarding the data, scripts, or methods, please contact:

Jean-Christophe Mouren & Benoit Ballester
Aix Marseille University, INSERM, TAGC, UMR 1090

Emails:

• jean-christophe.mouren@inserm.fr
• benoit.ballester@inserm.fr