Data from: Increased power to dissect adaptive traits in global sorghum diversity using a nested association mapping population

Adaptation of domesticated species to diverse agroclimatic regions has led to abundant trait diversity. However, the resulting population structure and genetic heterogeneity confounds association mapping of adaptive traits. To address this challenge in sorghum [Sorghum bicolor (L.) Moench]—a widely adapted cereal crop—we developed a nested association mapping (NAM) population using 10 diverse global lines crossed with an elite reference line RTx430. We characterized the population of 2214 recombinant inbred lines at 90,000 SNPs using genotyping-by-sequencing. The population captures ∼70% of known global SNP variation in sorghum, and 57,411 recombination events. Notably, recombination events were four- to fivefold enriched in coding sequences and 5′ untranslated regions of genes. To test the power of the NAM population for trait dissection, we conducted joint linkage mapping for two major adaptive traits, flowering time and plant height. We precisely mapped several known genes for these two traits, and identified several additional QTL. Considering all SNPs simultaneously, genetic variation accounted for 65% of flowering time variance and 75% of plant height variance. Further, we directly compared NAM to genome-wide association mapping (using panels of the same size) and found that flowering time and plant height QTL were more consistently identified with the NAM population. Finally, for simulated QTL under strong selection in diversity panels, the power of QTL detection was up to three times greater for NAM vs. association mapping with a diverse panel. These findings validate the NAM resource for trait mapping in sorghum, and demonstrate the value of NAM for dissection of adaptive traits.

驯化物种对多样农业气候区域的适应，孕育了丰富的性状多样性。然而由此形成的种群结构与遗传异质性，会干扰适应性性状的关联作图分析。针对广泛适配的谷类作物高粱（*Sorghum bicolor* (L.) Moench）中的这一难题，我们以10份全球多样化种质与优良参考系RTx430杂交，构建了嵌套关联作图（Nested Association Mapping, NAM）群体。我们利用测序分型（Genotyping-by-Sequencing, GBS）技术，对由2214份重组自交系（recombinant inbred lines, RILs）构成的NAM群体开展了9万个单核苷酸多态性（Single Nucleotide Polymorphism, SNPs）位点的基因分型。该群体覆盖了全球已知高粱SNP变异的约70%，并涵盖57411个重组事件。值得注意的是，重组事件在基因编码序列与5'非翻译区出现了4至5倍的富集现象。为验证该NAM群体用于性状解析的效能，我们针对两个核心适应性性状——开花期与株高开展了联合连锁作图分析。我们精准定位了调控这两个性状的多个已知基因，并鉴定出若干新增的数量性状位点（Quantitative Trait Locus, QTL）。当同时纳入所有SNP位点进行分析时，遗传变异可解释65%的开花期表型变异与75%的株高表型变异。此外，我们直接将NAM群体与同规模的全基因组关联作图群体进行对比，结果显示，相较于全基因组关联作图，NAM群体能够更稳定地鉴定出开花期与株高相关QTL。最后，针对多样性群体中经历强选择的模拟QTL，NAM群体的QTL检测效能最高可达同规模多样性群体关联作图的3倍。上述研究结果验证了该NAM资源在高粱性状定位中的应用价值，并证实了NAM群体在适应性性状解析中的重要意义。