Analysis of Brugada syndrome loci reveals that fine-mapping clustered GWAS hits enhances the annotation of disease-relevant variants

Mel·Lina Pinsach-Abuin,Bernat Del Olmo,Adrian Pérez-Agustin,Jesus Mates,Catarina Allegue,Anna Iglesias,Qi Ma,Daria Merkurjev,Sergiy Konovalov,Jing Zhang,Farah Sheikh,Amalio Telenti,Josep Brugada,Ramon Brugada,Melissa Gymrek,Julia Di Iulio,Ivan Garcia-Bassets,Sara Pagans

doi:10.1016/j.xcrm.2021.100250

Abstract

SummaryGenome-wide association studies (GWASs) are instrumental in identifying loci harboring common single-nucleotide variants (SNVs) that affect human traits and diseases. GWAS hits emerge in clusters, but the focus is often on the most significant hit in each trait- or disease-associated locus. The remaining hits represent SNVs in linkage disequilibrium (LD) and are considered redundant and thus frequently marginally reported or exploited. Here, we interrogate the value of integrating the full set of GWAS hits in a locus repeatedly associated with cardiac conduction traits and arrhythmia, SCN5A-SCN10A. Our analysis reveals 5 common 7-SNV haplotypes (Hap1–5) with 2 combinations associated with life-threatening arrhythmia—Brugada syndrome (the risk Hap1/1 and protective Hap2/3 genotypes). Hap1 and Hap2 share 3 SNVs; thus, this analysis suggests that assuming redundancy among clustered GWAS hits can lead to confounding disease-risk associations and supports the need to deconstruct GWAS data in the context of haplotype composition.

Highlights

For over a decade, genome-wide association studies (GWASs) have laid the foundation for identifying heritable traits and disease-associated features.[1]
The observation of hits in clusters fits with a model of ‘‘haplotype blocks,’’ or genomic regions inherited as single sets across generations with internal singlenucleotide variants (SNVs) transmitted in linkage disequilibrium (LD).[3,4,5,6,7,8,9,10]
As each block-partitioning method relies on different LD principles,[40] some of these predictions were discordant

Summary

Introduction

Genome-wide association studies (GWASs) have laid the foundation for identifying heritable traits and disease-associated features.[1] GWASs compare common singlenucleotide variants (SNVs) between populations differing in a phenotypic trait—often a clinical symptom—to identify trait/disease-associated variants. The strength of the individual associations is not a readout of functionality and can vary across populations.[2] Despite this fact, the GWAS field has historically focused on the lead SNVs and has inferred redundancy on the remaining associations, often reported in supplemental information for this reason. The observation of hits in clusters fits with a model of ‘‘haplotype blocks,’’ or genomic regions inherited as single sets (called haplotypes) across generations with internal SNVs transmitted in linkage disequilibrium (LD).[3,4,5,6,7,8,9,10]

Objectives

Methods

Results

Discussion

Conclusion