Premium
Multitrait genome‐wide association analysis of Populus trichocarpa identifies key polymorphisms controlling morphological and physiological traits
Author(s) -
Chhetri Hari B.,
MacayaSanz David,
Kainer David,
Biswal Ajaya K.,
Evans Luke M.,
Chen JinGui,
Collins Cassandra,
Hunt Kimberly,
Mohanty Sushree S.,
Rosenstiel Todd,
Ryno David,
Winkeler Kim,
Yang Xiaohan,
Jacobson Daniel,
Mohnen Debra,
Muchero Wellington,
Strauss Steven H.,
Tschaplinski Timothy J.,
Tuskan Gerald A.,
DiFazio Stephen P.
Publication year - 2019
Publication title -
new phytologist
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.742
H-Index - 244
eISSN - 1469-8137
pISSN - 0028-646X
DOI - 10.1111/nph.15777
Subject(s) - biology , genome wide association study , single nucleotide polymorphism , genetic association , genetics , genetic architecture , quantitative trait locus , heritability , populus trichocarpa , candidate gene , association mapping , gene , phenotype , snp , missing heritability problem , trait , expression quantitative trait loci , genome , genotype , computer science , programming language
Summary Genome‐wide association studies (GWAS) have great promise for identifying the loci that contribute to adaptive variation, but the complex genetic architecture of many quantitative traits presents a substantial challenge. We measured 14 morphological and physiological traits and identified single nucleotide polymorphism (SNP)‐phenotype associations in a Populus trichocarpa population distributed from California, USA to British Columbia, Canada. We used whole‐genome resequencing data of 882 trees with more than 6.78 million SNPs, coupled with multitrait association to detect polymorphisms with potentially pleiotropic effects. Candidate genes were validated with functional data. Broad‐sense heritability ( H 2 ) ranged from 0.30 to 0.56 for morphological traits and 0.08 to 0.36 for physiological traits. In total, 4 and 20 gene models were detected using the single‐trait and multitrait association methods, respectively. Several of these associations were corroborated by additional lines of evidence, including co‐expression networks, metabolite analyses, and direct confirmation of gene function through RNAi. Multitrait association identified many more significant associations than single‐trait association, potentially revealing pleiotropic effects of individual genes. This approach can be particularly useful for challenging physiological traits such as water‐use efficiency or complex traits such as leaf morphology, for which we were able to identify credible candidate genes by combining multitrait association with gene co‐expression and co‐methylation data.