Analysis code for the ENCODE-rE2G paper

This repository contains code used for analyses presented in the ENCODE-rE2G ENCODE companion paper, as well as links to the used benchmarking pipelines.

Using the ENCODE-rE2G model

To apply pre-trained ENCODE-rE2G models or train new models, go to the ENCODE-rE2G model repository found here: https://github.com/EngreitzLab/ENCODE_rE2G.

Setting up the analysis code repository

To obtain a local copy of the repository, including the crispr_analyses/CRISPR_comparison submodule containing the CRISPR benchmarking pipeline version used for analyses in this paper, open a terminal and run:

git clone --recurse-submodules git@github.com:EngreitzLab/ENCODE-rE2G-Paper.git

This should take less than a minute.

Analysis code

The following subdirectories contain code used to perform analyses shown in the paper:

crispr_analyses:

Snakemake workflow for CRISPR benchmarking and other related analyses results shown in Main Figures 1, 2, 4 and 5, Extended Data Figures 1, 2, 3, 7, Supplementary Figures S1.1, S1.2, S7, S8, S9, S10, as well as Supplementary Tables S2, S9, S10, S14, S15, S16.

gwas_analyses:

Code used to compute GWAS metrics shown in analyses of the manuscript. GWAS benchmarks were performed using the GWAS benchmarking pipeline (see below).

process_dnase_metadata:

Code to process DNase-seq metadata from the ENCODE portal to select input files to generate ENCODE-rE2G predictions for 1,458 human DNase-seq experiments.

other analyses:

This subdirectory contains code used in other analyses across the manuscript, including annotating CRISPR data with chromatin categories, analyzing enhancer synergy and analyses related to 3D contact, enhancer-gene correlation and promoter classes.

Benchmarking pipelines

Following pipelines were used to benchmark enhancer-gene predictions against CRISPR, eQTL and GWAS data:

CRISPR benchmarks

A copy of the CRISPR benchmarking used to intersect predictions with CRISPR enhancer perturbation results can be found in the crispr_analyses subdirectory or in the original GitHib repository for general use: https://github.com/EngreitzLab/CRISPR_comparison.

eQTL benchmarks

All eQTL benchmarking analyses, including figures shown in the paper, were computed using the eQTL benchmarking pipeline available here: https://github.com/EngreitzLab/eQTLEnrichment.

GWAS benchmarks

The GWAS benchmarks were performed using the GWAS benchmarking pipeline available here: https://github.com/Deylab999MSKCC/e2g-benchmarking. Additional analysis code for GWAS-related analyses is available in the gwas_analyses subdirectory.