背景回顾：Genome-wide association studies (GWAS) identify loci underlying the variation of complex traits. One of the main limitations of GWAS is the availability of reliable phenotypic data, particularly for long-lived tree species. 

提出问题：Although an extensive amount of phenotypic data already exists in breeding programs, accounting for its high heterogeneity is a great challenge.

主要研究：We combine spatial and factor-analytics analyses to standardize the heterogeneous data from 120 field experiments of 483,424 progenies of Norway spruce to implement the largest reported GWAS for trees using 134 605 SNPs from exome sequencing of 5056 parental trees.

结果：We identify 55 novel quantitative trait loci (QTLs) that are associated with phenotypic variation. The largest number of QTLs is associated with the budburst stage, followed by diameter at breast height, wood quality, and frost damage. Two QTLs with the largest effect have a pleiotropic effect for budburst stage, frost damage, and diameter and are associated with MAP3K genes. Genotype data called from exome capture, recently developed SNP array and gene expression data indirectly support this discovery. Several important QTLs associated with growth and frost damage have been verified in several southern and northern progeny plantations, indicating that these loci can be used in QTL-assisted genomic selection. 

结论：Our study also demonstrates that existing heterogeneous phenotypic data from breeding programs, collected over several decades, is an important source for GWAS and that such integration into GWAS should be a major area of inquiry in the future.

 摘 要 

全基因组关联分析（GWAS）能够鉴定复杂性状背后的潜在遗传变异基础。而GWAS的一个主要限制性因素是可靠的表型数据，对于多年生的林木而言尤其如此。尽管在育种过程中，积累了大量的表型数据，但如何解释其高度异质性是一个巨大的挑战。本文中，作者结合了结合空间和因子分析来标准化483424个后代的120个田间试验的异质数据，并基于对5056个母树的外显子测序所获得的134 605个SNP，对挪威云杉以实施规模最大的树木GWAS研究。作者鉴定了55个与表型变异相关的新的数量性状位点（QTL）。最多的QTLs与爆芽期有关，其次是与胸径、木材质量和冻害等性状相关。两个效应最大的QTLs对爆芽期、冻害和胸径具有多效性，并与MAP3K基因相关。来自外显子测序所捕获的基因型数据、最近开发的SNP矩阵和基因表达数据都间接支持了本文的发现。一些重要的QTLs与生长和冻害相关，并且在一些南方和北方的子代人工林中得到验证，表明这些位点可用于QTL辅助的基因组选择。本文的研究还表明，几十年来从育种项目收集的现有异质表型数据是GWAS研究项目中所需表型数据的一个重要来源，将这些数据整合进GWAS研究中应该是未来研究的一个主要领域。