|LI, CHUNHUI - Chinese Academy Of Agricultural Sciences|
|LI, YONG-XIANG - Chinese Academy Of Agricultural Sciences|
|WU, XUN - Chinese Academy Of Agricultural Sciences|
|SHI, YUNSU - Chinese Academy Of Agricultural Sciences|
|SONG, YANCHUN - Chinese Academy Of Agricultural Sciences|
|ZHANG, DENGFENG - Chinese Academy Of Agricultural Sciences|
|RODGERS-MELNICK, ELI - Cornell University|
|Buckler, Edward - Ed|
|ZHANG, ZHIWU - Washington State University|
|LI, YU - Chinese Academy Of Agricultural Sciences|
|WANG, TIANYU - Chinese Academy Of Agricultural Sciences|
Submitted to: BMC Plant Biology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/2/2015
Publication Date: 9/21/2015
Citation: Li, C., Li, Y., Bradbury, P., Wu, X., Shi, Y., Song, Y., Zhang, D., Rodgers-Melnick, E., Buckler IV, E.S., Zhang, Z., Li, Y., Wang, T. 2015. Construction of high-quality recombination maps with low-coverage genomic sequencing for joint linkage analysis in maize. Biomed Central (BMC) Plant Biology. 13:78.
Interpretive Summary: Linkage mapping identifies regions of chromosomes that contain genetic variants that cause changes in observable traits. The regions controlling trait variation are called quantitative trait loci (QTL). In this study, one million single nucleotide polymorphisms (SNPs), which are used as genetic markers, and 7000 maize inbred lines were used to find QTL that control flowering time. This study combined two large nested association mapping (NAM) populations, the US-NAM, developed earlier and consisting of 25 families of inbreds each derived from two inbred parents, and the CN-NAM, first reported here, consisting of 11 families of lines bred from parents adapted to China. Flowering time data was collected in three locations in each of two years in China. A joint analysis of both NAMs found 37 significant QTL. Separate analysis found 29 and 18 QTL for the US and CN NAM populations, respectively. The paper compares QTL positions to known flowering time genes and showed that one quarter of the QTL contained one of the known genes.
Technical Abstract: A genome-wide association study (GWAS) is the foremost strategy used for finding genes that control human diseases and agriculturally important traits, but it often reports false positives. In contrast, its complementary method, linkage analysis, provides direct genetic confirmation, but with limited resolution. A joint approach, using multiple linkage populations, dramatically improves resolution and statistical power. For example, this approach has been used to confirm that many complex traits, such as flowering time controlling adaptation in maize, are controlled by multiple genes with small effects. In addition, genotyping by sequencing (GBS) at low coverage not only produces genotyping errors, but also results in large datasets, making the use of high-throughput sequencing technologies computationally inefficient or unfeasible. In this study, we converted raw SNPs into effective recombination bins. The reduced bins not only retain the original information, but also correct sequencing errors from low-coverage genomic sequencing. To further increase the statistical power and resolution, we merged a new temperate maize nested association mapping (NAM) population derived in China (CN-NAM) with the existing maize NAM population developed in the US (US-NAM). Together, the two populations contain 36 families and 7,000 recombinant inbred lines (RILs). One million SNPs were generated for all the RILs with GBS at low coverage. We developed high-quality recombination maps for each NAM population to correct genotyping errors and improve the computational efficiency of the joint linkage analysis. The original one million SNPs were reduced to 4,932 and 5,296 recombination bins with average interval distances of 0.34 cM and 0.28 cM for CN-NAM and US-NAM, respectively. The quantitative trait locus (QTL) mapping for flowering time (days to tasseling) indicated that the high-density, recombination bin map improved resolution of QTL mapping by 50 % compared with that using a medium-density map. We also demonstrated that combining the CN-NAM and US-NAM populations improves the power to detect QTL by 50 % compared to single NAM population mapping. Among the QTLs mapped by joint usage of the US-NAM and CN-NAM maps, 25 % of the QTLs overlapped with known flowering-time genes in maize. This study provides directions and resources for the research community, especially maize researchers, for future studies using the recombination bin strategy for joint linkage analysis. Available resources include efficient usage of low-coverage genomic sequencing, detailed positions for genes controlling maize flowering, and recombination bin maps and flowering- time data for both CN and US NAMs. Maize researchers even have the opportunity to grow both CN and US NAM populations to study the traits of their interest, as the seeds of both NAM populations are available from the seed repository in China and the US.