Soybean Research Institute, Nanjing Agricultural University, Nanjing, 210095, Jiangsu, China.
National Center for Soybean Improvement, Ministry of Agriculture, Nanjing, 210095, Jiangsu, China.
Theor Appl Genet. 2018 Dec;131(12):2581-2599. doi: 10.1007/s00122-018-3174-7. Epub 2018 Aug 30.
Eighty-six R1 QTLs accounting for 89.92% phenotypic variance in a soybean RIL population were identified using RTM-GWAS with SNPLDB marker which performed superior over CIM and MLM-GWAS with BIN/SNPLDB marker. A population (NJRIKY) composed of 427 recombinant inbred lines (RILs) derived from Kefeng-1 × NN1138-2 (MGII × MGV, MG maturity group) was applied for detecting flowering date (R1) quantitative trait locus (QTL) system in soybean. From a low-depth re-sequencing (~ 0.75 ×), 576,874 SNPs were detected and organized into 4737 BINs (recombination breakpoint determinations) and 3683 SNP linkage disequilibrium blocks (SNPLDBs), respectively. Using the association mapping procedures "Restricted Two-stage Multi-locus Genome-wide Association Study" (RTM-GWAS), "Mixed Linear Model Genome-wide Association Study" (MLM-GWAS) and the linkage mapping procedure "Composite Interval Mapping" (CIM), 67, 36 and 10 BIN-QTLs and 86, 14 and 23 SNPLDB-QTLs were detected with their phenotypic variance explained (PVE) 88.70-89.92% (within heritability 98.2%), 146.41-353.62% (overflowing) and 88.29-172.34% (overflowing), respectively. The RTM-GWAS with SNPLDBs which showed to be more efficient and reasonable than the others was used to identify the R1 QTL system in NJRIKY. The detected 86 SNPLDB-QTLs with their PVE from 0.02 to 30.66% in a total of 89.92% covered 51 out of 104 R1 QTLs in 18 crosses in SoyBase and 26 out of 139 QTLs in a nested association mapping population, while the rest 29 QTLs were novel ones. From the QTL system, 52 candidate genes were annotated, including the verified gene E1, E2, E9 and J, and grouped into 3 categories of biological processes, among which 24 genes were enriched into three protein-protein interaction networks, suggesting gene networks working together. Since NJRIKY involves only MGII and MGV, the QTL/gene system among MG000-MGX should be explored further.
利用基于 SNPLDB 标记的 RTM-GWAS 方法在大豆 RIL 群体中鉴定了 86 个 R1 QTL，这些 QTL 解释了 89.92%的表型变异，优于基于 BIN/SNPLDB 标记的 CIM 和 MLM-GWAS 方法。使用由 Kefeng-1 × NN1138-2 (MGII × MGV，MG 成熟组) 衍生的 427 个重组自交系 (RIL) 组成的群体 (NJRIKY) 检测大豆开花日期 (R1) 数量性状位点 (QTL) 系统。通过低深度重测序 (~0.75×)，检测到 576874 个 SNP，并分别组织成 4737 个 BIN (重组断点确定) 和 3683 个 SNP 连锁不平衡块 (SNPLDB)。利用关联作图程序“受限两阶段多基因全基因组关联研究”(RTM-GWAS)、“混合线性模型全基因组关联研究”(MLM-GWAS)和连锁作图程序“复合区间作图”(CIM)，检测到 67、36 和 10 个 BIN-QTL 和 86、14 和 23 个 SNPLDB-QTL，其表型方差解释率 (PVE) 分别为 88.70-89.92% (遗传力内 98.2%)、146.41-353.62% (溢出) 和 88.29-172.34% (溢出)。与其他方法相比，基于 SNPLDB 的 RTM-GWAS 显示出更高的效率和合理性，用于鉴定 NJRIKY 中的 R1 QTL 系统。在 SoyBase 中 18 个杂交组合的 104 个 R1 QTL 中，检测到的 86 个 SNPLDB-QTL 中的 51 个具有 0.02-30.66%的 PVE，在嵌套关联作图群体中的 139 个 QTL 中，26 个具有 88.29-172.34%的 PVE，其余 29 个是新的 QTL。从 QTL 系统中，注释了 52 个候选基因，包括已验证的基因 E1、E2、E9 和 J，并分为 3 类生物过程，其中 24 个基因富集到三个蛋白质-蛋白质相互作用网络中，表明基因网络协同工作。由于 NJRIKY 仅涉及 MGII 和 MGV，因此应进一步探索 MG000-MGX 之间的 QTL/基因系统。
