玉米株高和穗位高QTL定位和全基因组选择探究

以郑58和B73为亲本构建的F₂群体，基于玉米48k液相探针捕获技术对株高和穗位高进行QTL定位和全基因组选择探究。结果表明，检测到20个株高的QTL，单个QTL的表型贡献率为2.13%～31.67%；检测到18个穗位高的QTL，单个QTL的表型贡献率为1.20%～27.49%，在4号染色体上的bin4.09区间发现存在1个同时控制株高和穗位高的主效QTL。通过改变训练群体大小和标记数目，用R语言软件的rrBLUP包对株高和穗位高进行全基因组选择，结果发现，株高和穗位高的预测精度分别为0.79和0.76。预测精度随着训练群体大小的增大而增加，当训练群体大小为50%时即可获得相对较高的预测精度。预测精度会随着标记数目的增加而增大，当标记数目为500时即可获得相对较高的预测精度。

QTL Mapping and Genomic Selection for Plant Height and Ear Height in Maize

QTL mapping and genomic selection were conducted in an F₂ population constructed from Zheng 58 and B73 based on maize 48K liquid phase probe capture technique. Twenty QTL of plant height were detected, and the phenotypic variation explained(PVE)by a single QTL ranged from 2.13%-31.67%. Eighteen QTL relate to ear height were detected, and the PVE of a single QTL ranged from 1.20% to 27.49%. Moreover, a major QTL controlling both plant height and ear height was found in bin4.09 on chromosome 4. The genomic selection of plant height and ear height was carried out with the rrBLUP package of R software. The prediction accuracies of genomic selection for plant height and ear height were 0.79 and 0.76, respectively. The prediction accuracy would increase as the training population size and marker density increase. When the training population size is 50%, a relatively high prediction accuracy could be obtained. When the number of markers is 500, the relatively high prediction accuracy can be obtained.