Regulatory Annotations Comparisons

This repository contains code to reproduce results of the manuscript: Cell specificty of regulatory annotations and their genetics effects on gene expression Processed datasets and a static version of the code have been submitted to Zenodo, DOI: 10.5281/zenodo.1413623

Each directory contains code for analyses specific to different figures of the paper. Everything is run using snakemake. The analysis directories follow this general pattern:

├── README.md
├── config.yaml : Cluster config specifications (if analysis is recommended to be run on a cluster)
├── scripts : Scripts for analyses
│   ├── script1.py
│   └── script2.R
└── Snakefile : To run the workflow

Requirements

The analyses use the following software:

• Python 3.6.3
• Snakemake 5.1.4
• R 3.3.2
• bedtools 2.26.0
• GREGOR 1.2.1
• GAT 1.3.5
• PLINK 1.9
• vcftools 1.15

To setup these pre-requisites, use the Anaconda/Miniconda Python3 distribution. The Conda package manager is used to obtain and deploy the defined software packages in the specified versions. These instructions are for the Linux platform

Step 1: Install Anaconda3

Assuming that you have a 64-bit system, on Linux, download and install Anaconda 3

$ wget https://repo.continuum.io/archive/Anaconda3-5.0.1-Linux-x86_64.sh
$ bash Anaconda3-5.0.1-Linux-x86_64.sh

Answer yes to the user agreement; go with the default installation path or specify your own. Answer yes to prepend the install location to your PATH.

Required for analyses computing enrichment of eQTL and GWAS in regulatory annotations, please install GREGOR manually. Edit the path to the GREGOR.pl script file in Snakefile_config

Step 2: Prepare analysis directory

Create a new directory and change into it. Clone this repository. This will produce the directory structure to replicate specific analyses. Edit the BASE_PATH in the Snakefile_config

Step 3: Create and activate environment

Create an environment named regulatory_comparisons with the required software using the environment.yaml file and activate it

$ conda env create --name regulatory_comparisons --file environment.yaml
$ source activate regulatory_comparisons

Now you can use the installed tools. Check if all went fine:

$ snakemake --help

Step 4: Execute analyses

Now change into the analysis directories and run snakemake using the respective Snakefiles. Each Snakefile will check for and download required data if missing as per the Snakefile_config in the top directory.

Below are the analysis directories corresponding to each figure:

├── Fig 2A-E: `summaryStatistics` :: `Snakefile`
│   └──	Supplementary Fig S1 :: `Snakefile_GAT`		
├── Fig 3: `histone_info_content` :: `Snakefile`
│   └──	Supplementary Fig S3, S4 :: `Snakefile`			
├── Fig 4A: `enrichment_distanceToNearestGeneTSS` :: `Snakefile_binBylclESI`
│   ├── Supplementary Fig S6 :: `Snakefile_allPCgenes`	
│   ├── Supplementary Fig S8: Automatically runs subworkflow in `lclESI_GTExV7`
│   └──	Supplementary Fig S9 :: `Snakefile_binBylclESI`
├── Fig 4B: `enrichment_eQTL` :: `Snakefile_binByESI`
│   ├── Supplementary Fig S7 :: `Snakefile_bulkEnrichment`	
│   ├── Automatically runs subworkflow in `lclESI_GTExV7`
│   └──	Supplementary Fig S10 :: `Snakefile_binByESI`	
├── Fig 5A, B: `effectSizeDistribution_bloodEqtl`
│   └── Fig 5A, B, Supplementary Fig S11, table 1 :: `Snakefile`
├── Fig 5C: `effectSizeDistribution_dsqtl`
│   └── Fig 5C :: `Snakefile_dsQTLEffect`
├── Fig 5D: `effectSizeDistribution_dsqtl`
│   └── Fig 5D, Supplementary Fig S13 :: `Snakefile_allelicBiasEffect`
├── Supplementary Fig S2: `coverageChromatinStates` :: `Snakefile`
├── Supplementary Fig S5, Table S2: `enrichment_gwas_extended` :: `Snakefile`
├── Supplementary Fig S12: `effectSizeDistribution_eqtl` :: `Snakefile`