Multimorbidity Atlas

To investigate trait or phenotype multimorbidity by integrating chromatin interaction and expression quantitative trait loci data.

Data

• GWAS Catalog all associations v1.0.1 from here
• Rao et al (2014)'s Hi-C libraries of GM12878, IMR90, HMEC, NHEK, K562, HUVEC, and KBM7 cell lines from the GEO repository
• GTEx multi-tissue eQTLs data analysis v4
• OMIM's genemap2 data

Visualisation

The figures generated in this study (and their underlying code) can be viewed in visualisation.html or the visualisation.Rmd R Markdown file.

Getting Started

• To reproduce this work, clone this repository with:

  git clone https://github.com/Genome3d/multimorbidity-atlas.git
  

• Install the [CoDeS3D][https://github.com/alcamerone/codes3d] pipeline:

  cd scripts/python/
  git clone https://github.com/alcamerone/codes3d.git
  

• To quickly find multimorbid traits without running the entire methods, download and unzip the spatial eQTL data for traits into the results directory (1.59 GB). Then run step 9 in Method overview section below.

Method overview

1. Obtaining SNPs from GWAS Catalog associations file from the command line:

   cd scripts/python/
   ./extract_gwas_snps.py
   

2. Identification of spatial eQTLs from GWAS SNPs:

   cd ../bash/
   ./batch_codes3d.sh
   

3. Saving eQTLs to database:

   cd ../python/
   ./init_db.py
   

4. Get eQTLs associated with complex traits:

   ./get_trait_eqtls.py
   

5. Obtain significant tissue-specific spatial eQTLs:

   cd ../bash/
   ./produce_summaries.sh
   

6. Get eQTL-eGene interactions in tissues

   cd ../python/
   ./get_eqtl_interactions.py
   

7. Construct matrices of shared eQTLs, eGenes ratios among complex traits. Plus control analysis:

   ./control_analysis.py
   

8. Cluster complex traits by shared eGenes and eQTLs

   cd ../R/
   ./convex_biclustering.R
   

9. Find traits that share eGenes. Run command with -h for help:

   cd ../python/
   ./get_cluster.py