Quantitative trait loci identification and meta-analysis for rice panicle-related traits

Rice yield is a complex trait controlled by quantitative trait loci (QTLs). In the past three decades, thousands of QTLs for rice yield traits have been detected, but only a very small percentage has been cloned to date, as identifying the QTL genes requires a substantial investment of time and money. Meta-analysis provides a simple, reliable, and economical method for integrating information from multiple QTL studies across various environmental and genetic backgrounds, detecting consistent QTLs powerfully and estimating their genetic positions precisely. In this study, we aimed to locate consistent QTL regions associated with rice panicle traits by applying a genome-wide QTL meta-analysis approach. We first conducted a QTL analysis of 5 rice panicle traits using 172 plants in 2011 and 138 plants in 2012 from an F-2 population derived from a cross between Nipponbare and H71D rice cultivators. A total of 54 QTLs were detected, and these were combined with 1085 QTLs collected from 82 previous studies to perform a meta-analysis using BioMercator v4.2. The integration of 82 maps resulted in a consensus map with 6970 markers and a total map length of 1823.1 centimorgan (cM), on which 837 QTLs were projected. These QTLs were then integrated into 87 meta-quantitative trait loci (MQTLs) by meta-analysis, and the 95 % confidence intervals (CI) of them were smaller than the mean value of the original QTLs. Also, 30 MQTLs covered 47 of the 54 QTLs detected from the cross between Nipponbare and H71D in this study. Among them, the two major and stable QTLs, spp10.1 and sd10.1, were found to be included in MQTL10.4. The three other major QTLs, pl3.1, sb2.1, and sb10.1, were included in MQTL3.3, MQTL2.2, and MQTL10.3, respectively. A total of 21 of the 87 MQTLs' phenotypic variation were > 20 %. In total, 24 candidate genes were found in 15 MQTLs that spanned physical intervals < 0.2 Mb, including genes that have been cloned previously, e.g., EP3, LP, MIP1, HTD1, DSH1, and OsPNH1. However, it would be beneficial to identify a greater number of candidate genes from these MQTLs. Mining new genes that modulate yield and its related traits would assist researchers to better understand the relevant molecular mechanisms. The MQTLs found in this study that have small physical and genetic intervals are useful not only for marker-assisted selection and pyramiding, but they also provide important information of rice yield and related gene mining for future research.