enhancer3D - Help

This page provides an overview of the assumptions, input data, and processing steps used to build the enhancer3D database. For details on the 3D chromatin structure modeling algorithm employed (3D-GNOME 3.0), please refer to the main Help and Tutorial pages of the 3D-GNOME server.

This section also outlines the types of data available for download in the Download Section.

Introduction

enhancer3D is a public database of 3D chromatin structures and enhancer–promoter (E–P) distance profiles for both archaic and modern human genomes. It allows comparative analysis of chromatin architecture, focusing on how structural variations may have altered regulatory interactions during human evolution.

The database offers two core analysis modules:

All 3D models were generated using the GPU-accelerated cudaMMC engine, as used in the 3D-GNOME 3.0 platform, based on high-confidence CTCF interaction data. The results are accessible via an interactive viewer with additional downloadable 3D models in .cif format.

Users can explore 3D chromatin topology, visualize enhancer–promoter distances, and retrieve model data tailored to their region or gene of interest.

For technical details on the 3D modeling approach, visit the main 3D-GNOME Help page. To download datasets from enhancer3D, see the Download section.

Data Description and Modeling Pipeline

The enhancer3D database was constructed through a systematic modeling pipeline built upon the 3D-GNOME 3.0 platform. Structural variants (SVs) for archaic human genomes — Denisovans and Neanderthals — were obtained from the Simons Genome Diversity Project, converted from hg19 to hg38 (liftover), and filtered to retain only autosomal, bi-allelic deletions and duplications.

The chromatin interaction data used for modeling comes from the 4D Nucleome (4DN) Consortium. For three modern human cell lines — GM12878, H1ESC, and HFFC6 — we extracted high-confidence CTCF-mediated chromatin interactions, specifically PET clusters with ≥3 tags (3+), as input for modeling.

For each structural variant, a ±1 Mb window was defined around the affected region. Within these windows, the CTCF interaction networks of each cell line were perturbed using the approach described in Sadowski et al., Genome Biology 2019, producing predicted chromatin 3D interactions for Denisovan and Neanderthal genomes.

These interaction maps — original and SV-modified — were then input into the GPU-accelerated cudaMMC engine used in the 3D-GNOME 3.0 platform to generate ensembles of 100 3D chromatin models per region for each genome (modern, Denisovan, and Neanderthal).

Genomic annotations were mapped onto these models: gene positions were taken from GENCODE v40 (hg38) and enhancer coordinates from EnhancerAtlas 2.0 (liftover from hg19 to hg38). Using this combined data, the 3D distances between enhancers and promoters were calculated, distributions were compared, and mean distance shifts between modern and archaic models were computed.

In the Download Section, users can access:

Beyond archaic vs modern human analysis, the enhancer3D browser also supports full-chromosome data exploration for modern cell lines. The following data can be downloaded for each of the three cell lines:

enhancer3D Request Form & Results – Tutorial

The enhancer3D request form allows users to explore 3D chromatin organization and enhancer–promoter architecture in two distinct modes:

Depending on the selected mode, you will be prompted to choose relevant cell lines. In archaic vs modern, an additional step involves selecting the archaic genome (Neanderthal or Denisovan) to compare against the chosen modern human cell line.

For cross-cell line mode, you can choose whether to analyze a specific genomic region or view an entire chromosome.

Next, provide either a genomic region in the format chr:start-end or specify an autosomal chromosome only (e.g., chr14). Alternatively, you can enter a gene name or Ensembl ID to retrieve its corresponding region automatically.

For convenience, a Load example button is available to prefill the form with sample values. Once ready, click Submit.

The platform will redirect to a results page where an IGV browser displays relevant data tracks, including:

For archaic vs modern analyses, a detailed table appears below the browser listing enhancer–promoter pairs found in the selected region. Each row contains genomic coordinates, enhancer scores, average 3D distances in both genomes, and statistical significance metrics. You can:

Additionally, a Open 3D view button below the browser opens a separate interactive page displaying 3D chromatin structures. In this viewer, genes are shown in red and enhancers in yellow, enabling visual exploration of spatial enhancer–promoter relationships.