This pipeline facilitates easy usage of coloc (Giambartolomei et al. 2014, Wallace 2020) with GWAS and eQTL data.

Colocalization analysis is used to detect genetic causality between two different GWAS traits.

Coloc-wrapper performs genetic colocalization analysis for GWAS and eQTL datasets in a given region using coloc.abf() function from Coloc R-package. It calculates posterior probabilities for the following five hypothesis for each gene in the region under the assumption of a single causal variant for each trait:

: no association
: association to trait 1 only
: association to trait 2 only
: association to both traits, distinct causal variants
: association to both traits, shared causal variant

The posterior probability of hypothesis 4, PP4, determines the possible colocalization. A common threshold for it is PP4 > 0.8.

To get started, look at this minimal example.

• R version >3.6.2
• R-packages: "coloc", "data.table", "ggplot2", "optparse", "R.utils"
• tabix

sudo docker build -t coloc-wrapper -f docker/Dockerfile .

sudo docker run -it -v /mnt/disks/1/projects/COLOC:/COLOC -w /COLOC coloc-wrapper /bin/bash

The input files are the following:

eQTL data can be found here: https://www.ebi.ac.uk/eqtl/

Running coloc-wrapper involves two steps:

1. Trimming data, both GWAS and eQTL, according to a predefined region
2. Running coloc

• file: GWAS or eQTL file path or url
• region: genomic region of interest, format chr:start-end
• out: output file

Rscript extdata/step1_subset_data.R	\
	--file=ftp://ftp.ebi.ac.uk/pub/databases/spot/eQTL/csv/Lepik_2017/ge/Lepik_2017_ge_blood.all.tsv.gz \
	--region="1:10565520-10965520" \
  	--out=tmp.txt

• gwas: GWAS summary statistics file for one region
• eqtl: eQTL summary statistics file for one region
• header_gwas: header of GWAS file, named vector in quotes
• header_eqtl: header of eQTL file, named vector in quotes
• info_gwas: options for GWAS dataset, more info here
  • type: the type of data in dataset - either "quant" or "cc" to denote quantitative or case-control
  • s: for a case control dataset, the proportion of samples in dataset that are cases
  • N: number of samples in the dataset
• info_eqtl: options for eQTL dataset, more info here
  • type: the type of data in dataset - either "quant" or "cc" to denote quantitative or case-control
  • sdY: for a quantitative trait, the population standard deviation of the trait. if not given, it can be estimated from the vectors of varbeta and MAF
  • N: number of samples in the dataset
• p1: the prior probability that any random SNP in the region is associated with exactly trait 1
• p2: the prior probability that any random SNP in the region is associated with exactly trait 2
• p12: the prior probability that any random SNP in the region is associated with both traits
• locuscompare_thresh: PP4 threshold that plots the locuscompare plots
• out: output file

Rscript extdata/step2_run_coloc.R	\
	--eqtl="extdata/Lepik_2017_ge_blood_chr1_ENSG00000142655_ENSG00000130940.all.tsv" \
	--gwas="extdata/I9_VARICVE_chr1.tsv.gz" \
	--header_eqtl="c(varid = 'rsid', pvalues = 'pvalue', MAF = 'maf', gene_id = 'gene_id')" \
	--header_gwas="c(varid = 'rsids', pvalues = 'pval', MAF = 'maf')" \
	--info_gwas="list(type = 'cc', s = 11006/117692, N  = 11006 + 117692)" \
	--info_eqtl="list(type = 'quant', sdY = 1, N = 491)" \
	--p1=1e-4 \
	--p2=1e-4 \
	--p12=5e-6 \
	--locuscompare_thresh=0.8 \
	--out="Coloc_example.txt" \

• gene_id: gene identifier
• nsnps: number of SNPs included in colocalization
• PP.H0.abf: Posterior probability that neither trait has a genetic association in the region
• PP.H1.abf: Posterior probability that only trait 1 has a genetic association in the region
• PP.H2.abf: Posterior probability that only trait 2 has a genetic association in the region
• PP.H3.abf: Posterior probability that both traits are associated, but with different causal variants
• PP.H4.abf: Posterior probability that both traits are associated and share a single causal variant

For more details to output columns see coloc-package.

An alternative coloc-wrapper: https://github.com/eQTL-Catalogue/colocalisation

Rscript -e 'testthat::test_dir("tests/testthat/")'

• Original coloc paper: Giambartolomei, C., Vukcevic, D., Schadt, E.E., Franke, L., Hingorani, A.D., Wallace, C., Plagnol, V., 2014. Bayesian Test for Colocalisation between Pairs of Genetic Association Studies Using Summary Statistics. PLOS Genetics 10, e1004383. https://doi.org/10.1371/journal.pgen.1004383
• Updates on coloc: Wallace, C., 2020. Eliciting priors and relaxing the single causal variant assumption in colocalisation analyses. PLOS Genetics 16, e1008720. https://doi.org/10.1371/journal.pgen.1008720
• Importance of visualizing locus: Liu, B., Gloudemans, M.J., Rao, A.S., Ingelsson, E., Montgomery, S.B., 2019. Abundant associations with gene expression complicate GWAS follow-up. Nat Genet 51, 768–769. https://doi.org/10.1038/s41588-019-0404-0 (see also locuscomparer)