Summary

You can find a preprint version of the manuscript
Ines Assum & Julia Krause et al., Tissue-specific multi-omics analysis of atrial fibrillation
here: https://doi.org/10.1101/2020.04.06.021527

Abstract:
Genome-wide association studies (GWAS) for atrial fibrillation (AF) have uncovered numerous disease-associated variants. Their underlying molecular mechanisms, especially consequences for mRNA and protein expression remain largely elusive. Thus, refined multi-omics approaches are needed for deciphering the underlying molecular networks. Here, we integrate genomics, transcriptomics, and proteomics of human atrial tissue in a cross-sectional study to identify widespread effects of genetic variants on both transcript (cis-eQTL) and protein (cis-pQTL) abundance. We further establish a novel targeted trans-QTL approach based on polygenic risk scores to determine candidates for AF core genes. Using this approach, we identify two trans-eQTLs and five trans-pQTLs for AF GWAS hits, and elucidate the role of the transcription factor NKX2-5 as a link between the GWAS SNP rs9481842 and AF. Altogether, we present an integrative multi-omics method to uncover trans-acting networks in small datasets and provide a rich resource of atrial tissue-specific regulatory variants for transcript and protein levels for cardiovascular disease gene prioritization.

We address a key hypothesis about the existence of core genes as postulated in the omnigenic model by Liu et al., Cell (2019). Core genes are central genes with trans-associations to GWAS loci, whose expression levels directly affect a disease phenotype. Here we sought to identify candidate core genes for AF to understand the contribution of trans-genetic effects in the pathology of AF. To prioritize genes satisfying the properties predicted by the omnigenic model, we evaluated the accumulation of trans-effects, their relevance in gene regulatory networks, and the disease association by the following strategy:

• We evaluated the cumulated trans-effects of AF-associated variants on expression by ranking genes based on their correlation of mRNA and protein abundance with the PRS for AF, so called expression/protein quantitative trait score (eQTS/pQTS, Võsa et al., bioRxiv 2018). While correcting for possible cis-effects by including the top SNP per independent cis-QTL loci, the PRS served as a proxy for an aggregation of AF-related trans-effects across the whole genome.
• To identify genes sharing molecular function and representing biological networks that propagate trans-effects to core genes, gene set enrichment analysis (GSEA) was performed on the eQTS and pQTS rankings. Genes driving the enrichment of multiple gene sets were selected as core gene candidates.
• The link between the core gene candidates and AF was established based on a significant trans-eQTL or pQTL for an AF GWAS hit and further supported by differential protein abundance analysis.

Data availability

Results derived using this code are available as supplementary material and in the Zenodo repository https://doi.org/10.5281/zenodo.5080229 .

Run your own analysis

Want to try our approach? Let's get started with a short tutorial! You can have a look at the html, run it in a R markdown or visit our notebook on google colab! You can run all scripts without having the R packages installed, but if you want to run your own data, you can use the google colab to install everything you need in less than 15 minutes! Note that for simplicity, no correction for cis-effects were performed in the tutorial.

Installation instructions

Analysis was done using R 3.4.1.

In general, four packages need to be installed

• devtools
• fgsea
• MatrixEQTL
• eQTLpipeline (optional)

install.packages("devtools")
BiocManager::install("fgsea", dependencies = T, clean = T)
library(devtools)
devtools::install_github("andreyshabalin/MatrixEQTL", force=T)
devtools::install_github("matthiasheinig/eQTLpipeline", force=T)

Additional packages are required to run the cis QTL pipeline with PEER analysis. For this, we provide a conda environment running on fedora 25, as PEER frequently causes problems when being installed as a R package. A docker will be supplied soon.
For installation, please

• create a new conda environment, using the yml-file conda env create -f r341peer.yml
• activate the environment conda activate r341peer
• start R and install packages MatrixEQTL and eQTLpipeline from github by running

library(devtools)
devtools::install_github("andreyshabalin/MatrixEQTL", force=T)
devtools::install_github("matthiasheinig/eQTLpipeline", force=T)

Tutorial

A short example of how to run our pre-selection approach with a short example dataset can be found as a R markdown, html document or as a google colab notebook.

If you want to run the analysis, in general the following R packages are required:

• devtools
• fgsea
• MatrixEQTL
• eQTLpipeline

Please also download gene set annotations here: GO biological processes

Cis-QTL analysis

Cis-QTL analyses were performed

You can find analysis code for the cis-QTL analysis in the folder qtl_pipeline

containing numerous scripts for

• preprocessing
• running PEER and then QTL analysis
• functional annotations
• comparison comparisons to other datasets (e.g. GTEx)
• colocalization adding colocalization analyses

Functional annotations

• build snp-gene pair annotations
• functions to evaluate enrichment of functional elements
• derivation of RBP binding sites
• NKX2-5 binding sites and TF activity

GWAS annotations

• calculate GWAS overlaps and enrichments

Trans-QTL analysis

Genome-wide polygenic scores for AF and CAD

Code for the computation of the polygenic risk score for AF on both our cohort and the 1000 genomes individuals can be found here.

AF candidate SNPs

The final list of 108 tested SNPs derived by pruning all SNPs annotated with AF in the GWAS catalog are supplied here.

eQTS/pQTS rankings and enrichments

GSEA is performed on eQTS/pQTS rankings and trans QTL analysis is carried out for the pruned AF SNPs with a subset of genes.

Power analysis

A power analysis was used to estimate the number of genes to consider in our analysis.

Replication of trans-results

Code for the replication analyses based on public data.

LICENSE

QTL approaches in human tissue for limited sample sizes Copyright (C) 2020 Ines Assum and Matthias Heinig

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see https://www.gnu.org/licenses/.