Genome-Wide Association Mapping and Genomic Selection for Alfalfa (Medicago sativa) Forage Quality Traits.
Genome-Wide Association Mapping and Genomic Selection for Alfalfa (Medicago sativa) Forage Quality Traits.
复制标题
苜蓿（Medicago sativa）草料质量特征的全基因组关联映射和基因组选择。
DOI：
10.1371/journal.pone.0169234
复制
发表时间：
2017
期刊：
影响因子：
3.7
通讯作者：
Annicchiarico P
中科院分区：
文献类型：
--
作者：
Biazzi E;Nazzicari N;Pecetti L;Brummer EC;Palmonari A;Tava A;Annicchiarico P
紫花苜蓿（Medicago sativa L.）是全球种植面积最大的饲料豆科植物，其饲料质量的遗传进展一直很差。本研究旨在探索基于亲本植物育种值的饲料品质性状的标记辅助选择（MAS）和基因组选择的机会。通过测序基因分型 (GBS) 对来自广泛参考群体的约 154 个基因型进行了基因分型，并对其密植半同胞后代在三种生长条件（夏季收获时，充分灌溉；夏季收获时，暂停灌溉；秋季收获时。由于种质×条件相互作用适度，对在不同条件下平均的后代值进行了性状标记分析。基因组选择利用了 11,450 个多态性 SNP 标记，而 8,494 个 M. truncatula 对齐标记的子集用于全基因组关联研究 (GWAS)。 GWAS 证实了品质性状的多基因控制，并且与表型相关性一致，表明茎和叶中给定性状的遗传控制存在显着不同的遗传控制。它在不同注释的基因中检测到几个与茎蛋白含量高度相关的单核苷酸多态性(SNP)。此外，它还发现了 8 号染色体上的一个小基因组区域，其中含有高浓度的与叶 ADL 相关的注释基因，其中包括一个可能参与木质素途径的基因。三种基因组选择模型，即 Ridge-regression BLUP、Bayes B 和 Bayesian Lasso，显示出相似的预测精度，而 SVR-lin 则不太准确。茎 NDFD 和叶蛋白质含量的准确度值为中等（0.3-0.4），叶 ADL 和 NDFD 的准确度适中，其他性状的准确度为低至极低。与之前对相同种质组的结果一起，本研究表明 GBS 数据可用于改善品质性状（通过基因组选择或 MAS）和饲料产量。
Genetic progress for forage quality has been poor in alfalfa (Medicago sativa L.), the most-grown forage legume worldwide. This study aimed at exploring opportunities for marker-assisted selection (MAS) and genomic selection of forage quality traits based on breeding values of parent plants. Some 154 genotypes from a broadly-based reference population were genotyped by genotyping-by-sequencing (GBS), and phenotyped for leaf-to-stem ratio, leaf and stem contents of protein, neutral detergent fiber (NDF) and acid detergent lignin (ADL), and leaf and stem NDF digestibility after 24 hours (NDFD), of their dense-planted half-sib progenies in three growing conditions (summer harvest, full irrigation; summer harvest, suspended irrigation; autumn harvest). Trait-marker analyses were performed on progeny values averaged over conditions, owing to modest germplasm × condition interaction. Genomic selection exploited 11,450 polymorphic SNP markers, whereas a subset of 8,494 M. truncatula-aligned markers were used for a genome-wide association study (GWAS). GWAS confirmed the polygenic control of quality traits and, in agreement with phenotypic correlations, indicated substantially different genetic control of a given trait in stems and leaves. It detected several SNPs in different annotated genes that were highly linked to stem protein content. Also, it identified a small genomic region on chromosome 8 with high concentration of annotated genes associated with leaf ADL, including one gene probably involved in the lignin pathway. Three genomic selection models, i.e., Ridge-regression BLUP, Bayes B and Bayesian Lasso, displayed similar prediction accuracy, whereas SVR-lin was less accurate. Accuracy values were moderate (0.3–0.4) for stem NDFD and leaf protein content, modest for leaf ADL and NDFD, and low to very low for the other traits. Along with previous results for the same germplasm set, this study indicates that GBS data can be exploited to improve both quality traits (by genomic selection or MAS) and forage yield.