Nontrivial Replication of Loci Detected by Multi-Trait Methods

Frontiers in Genetics

Yurii S. Aulchenko¹  Zhipeng Wang²  Xiao Feng³  Xia Shen⁴  Peter K. Joshi⁴  James F. Wilson⁴  Zheng Ning⁶  Yudi Pawitan⁶  Yakov A. Tsepilov⁸  Alexander K. Grishenko⁸  Sodbo Zh. Sharapov⁸  Masoud Shirali^1,10  Chris S. Haley^1,10 

[1] 0PolyOmica, 's-Hertogenbosch, Netherlands;Bioinformatics Center, Northeast Agricultural University, Harbin, China;Biostatistics Group, School of Life Sciences and School of Ecology, Sun Yat-sen University, Guangzhou, China;Centre for Global Health Research, Usher Institute, University of Edinburgh, Edinburgh, United Kingdom;College of Animal Science and Technology, Northeast Agricultural University, Harbin, China;Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden;Division of Biology, Novosibirsk State University, Novosibirsk, Russia;Institute of Cytology and Genetics SB RAS, Novosibirsk, Russia;Kurchatov Genomics Center, Institute of Cytology and Genetics SB RAS, Novosibirsk, Russia;MRC Human Genetics Unit, MRC Institute of Genetics and Molecular Medicine, University of Edinburgh, Western General Hospital, Edinburgh, United Kingdom;

关键词: pleiotropy;    multivariate analysis;    genome-wide association study;    cross-phenotype association;    replication;    genotype-phenotype map;   

DOI  :  10.3389/fgene.2021.627989

来源: DOAJ

【 摘 要 】

The ever-growing genome-wide association studies (GWAS) have revealed widespread pleiotropy. To exploit this, various methods that jointly consider associations of a genetic variant with multiple traits have been developed. Most efforts have been made concerning improving GWAS discovery power. However, how to replicate these discovered pleiotropic loci has yet to be discussed thoroughly. Unlike a single-trait scenario, multi-trait replication is not trivial considering the underlying genotype-multi-phenotype map of the associations. Here, we evaluate four methods for replicating multi-trait associations, corresponding to four levels of replication strength. Weak replication cannot justify pleiotropic genetic effects, whereas strong replication using our developed correlation methods can inform consistent pleiotropic genetic effects across the discovery and replication samples. We provide a protocol for replicating multi-trait genetic associations in practice. The described methods are implemented in the free and open-source R package MultiABEL.

【 授权许可】

Unknown