Combining meta-QTL with RNA-seq data to identify candidate genes of kernel row number trait in maize

Kernel row number is an important component of grain yield in maize. With development of classical quantitative trait loci (QTL) mapping and modern RNA-seq, numerous QTL and tissue-specific gene expression data were accumulated in previous studies. In this paper, a total of initial 373 QTL for grain yield (GY) and kernel row number (KRN) were collected based on 29 previous literatures. Fifty-four meta-QTL (MQTL) were detected via the meta-analysis method with IBM2 2008 Neighbors as a reference map, including 19 for GY and 35 for KRN. These MQTL were unevenly distributed on all 10 chromosomes. Chromosome 1 harbored the most initial QTL and MQTL, and chromosome 7 contained the least. Three MQTL for KRN have been overlapped with MQTL for GY on chromosomes 1 and 3. A total of 1,588 (46.07%) out of 3,447 genes located in the KRN MQTL regions were identified by gene expression data, and categorized into 101 significant GO terms. Meanwhile, six candidate genes were identified from MQTL regions, which are homologous to three functionally characterized genes found to participate in plant inflorescence development. The identified MQTL could be applied to marker-assisted selection (MAS) to facilitate yield architecture, QTL fine mapping and gene cloning in the maize community. Furthermore, the identified candidate genes could enhance the selection efficiency by MAS directly, and could illuminate molecular mechanisms of grain yield in maize