Construction and integration of genetic linkage maps from three multi-parent advanced generation inter-cross populations in rice

Sparse Phenotyping and Haplotype-Based Models for Genomic Prediction in Rice

The multi-environment genomic selection enables plant breeders to select varieties resilient to diverse environments or particularly adapted to specific environments, which holds a great potential to be used in rice breeding. To realize the multi-environment genomic selection, a robust training set with multi-environment phenotypic data is of necessity. Considering the huge potential of genomic prediction enhanced sparse phenotyping on the cost saving of multi-environment trials (MET), the establishment of a multi-environment training set could also benefit from it. Optimizing the genomic prediction methods is also crucial to enhance the multi-environment genomic selection. Using haplotype-based genomic prediction models is able to capture local epistatic effects which could be conserved and accumulated across generations much like additive effects thereby benefitting breeding. However, previous studies often used fixed length haplotypes composed by a few adjacent molecular markers disregarding the linkage disequilibrium (LD) which is of essential role in determining the haplotype length. In our study, based on three rice populations with different sizes and compositions, we investigated the usefulness and effectiveness of multi-environment training sets with varying phenotyping intensities and different haplotype-based genomic prediction models based on LD-derived haplotype blocks for two agronomic traits, i.e., days to heading (DTH) and plant height (PH). Results showed that phenotyping merely 30% records in multi-environment training set is able to provide a comparable prediction accuracy to high phenotyping intensities; the local epistatic effects are much likely existent in DTH; dividing the LD-derived haplotype blocks into small segments with two or three single nucleotide polymorphisms (SNPs) helps to maintain the predictive ability of haplotype-based models in large populations; modelling the covariances between environments improves genomic prediction accuracy. Our study provides means to improve the efficiency of multi-environment genomic selection in rice.

Genetic background- and environment-independent QTL and candidate gene identification of appearance quality in three MAGIC populations of rice

Many QTL have been identified for grain appearance quality by linkage analysis (LA) in bi-parental mapping populations and by genome-wide association study (GWAS) in natural populations in rice. However, few of the well characterized genes/QTL have been successfully applied in molecular rice breeding due to genetic background (GB) and environment effects on QTL expression and deficiency of favorable alleles. In this study, GWAS and LA were performed to identify QTL for five grain appearance quality-related traits using three multi-parent advanced generation inter-cross (MAGIC) populations. A total of 22 QTL on chromosomes 1-3, 5-8 were identified by GWAS for five traits in DC1, DC2 and 8way, and four combined populations DC12 (DC1+DC2), DC18 (DC1+8way), DC28 (DC2+8way) and DC128 (DC1+DC2+8way). And a total of 42 QTL were identified on all 12 chromosomes except 10 by LA in the three single populations. Among 20 QTL identified by GWAS in DC1, DC2 and 8way, 10, four and three QTL were commonly detected in DC18, DC28, and DC128, respectively. Similarly, among 42 QTL detected by LA in the three populations, four, one and two QTL were commonly detected in DC18, DC28, and DC128, respectively. There was no QTL mapped together in DC12 by both two mapping methods, indicating that GB could greatly affect the mapping results, and it was easier to map the common QTL among populations with similar GB. The 8way population was more powerful for QTL mapping than the DC1, DC2 and various combined populations. Compared with GWAS, LA can not only identify large-effect QTL, but also identify minor-effect ones. Among 11 QTL simultaneously detected by the two methods in different GBs and environments, eight QTL corresponded to known genes, including AqGL3b and AqGLWR3a for GL and GLWR, AqGW5a, AqGLWR5, AqDEC5 and AqPGWC5 for GW, GLWR, DEC and PGWC, and AqDEC6b and AqPGWC6b for DEC and PGWC, respectively. AqGL7, AqGL3c/AqGLWR3b, AqDEC6a/AqPGWC6a, and AqPGWC7 were newly identified and their candidate genes were analyzed and inferred. It was discussed to further improve grain appearance quality through designed QTL pyramiding strategy based on the stable QTL identified in the MAGIC populations.

Construction of Consensus Genetic Map With Applications in Gene Mapping of Wheat (Triticum aestivum L.) Using 90K SNP Array

Wheat is one of the most important cereal crops worldwide. A consensus map combines genetic information from multiple populations, providing an effective alternative to improve the genome coverage and marker density. In this study, we constructed a consensus map from three populations of recombinant inbred lines (RILs) of wheat using a 90K single nucleotide polymorphism (SNP) array. Phenotypic data on plant height (PH), spike length (SL), and thousand-kernel weight (TKW) was collected in six, four, and four environments in the three populations, and then used for quantitative trait locus (QTL) mapping. The mapping results obtained using the constructed consensus map were compared with previous results obtained using individual maps and previous studies on other populations. A simulation experiment was also conducted to assess the performance of QTL mapping with the consensus map. The constructed consensus map from the three populations spanned 4558.55 cM in length, with 25,667 SNPs, having high collinearity with physical map and individual maps. Based on the consensus map, 21, 27, and 19 stable QTLs were identified for PH, SL, and TKW, much more than those detected with individual maps. Four PH QTLs and six SL QTLs were likely to be novel. A putative gene called TraesCS4D02G076400 encoding gibberellin-regulated protein was identified to be the candidate gene for one major PH QTL located on 4DS, which may enrich genetic resources in wheat semi-dwarfing breeding. The simulation results indicated that the length of the confidence interval and standard errors of the QTLs detected using the consensus map were much smaller than those detected using individual maps. The consensus map constructed in this study provides the underlying genetic information for systematic mapping, comparison, and clustering of QTL, and gene discovery in wheat genetic study. The QTLs detected in this study had stable effects across environments and can be used to improve the wide adaptation of wheat cultivars through marker-assisted breeding.

QTL Mapping and Favorable Allele Mining of Nitrogen Deficiency Tolerance Using an Interconnected Breeding Population in Rice

Nitrogen is one of the most important nutrients for rice growth and development. Breeding of nitrogen deficiency tolerance (NDT) variety is considered to be the most economic measure to solve the constrain of low nitrogen stress on grain yield in rice. An interconnected breeding (IB) population of 497 lines developed using Huanghuazhan (HHZ) as the recurrent parent and eight elite lines as the donor parents were tested for five traits including grain yield, biomass, harvest index, thousand grain weight, and spikelet fertility under two nitrogen treatments in three growing seasons. Association analysis using 7,388 bins generated by sequencing identified a total of 14, 14, and 12 QTLs for the five traits under low nitrogen (LN), normal nitrogen (NN), and LN/NN conditions, respectively, across three seasons. Favorable alleles were dissected for the 40 QTLs at the 10 NDT regions, and OM1723 was considered as the most important parent with the highest frequency of favorable alleles contributing to NDT-related traits. Six superior lines all showed significantly higher GY in LN environments and similar GY under NN environments except for H10. Substitution mapping using near-isogenic introgression lines delimited the qTGW2-1, which was identified on chromosome 2 under LN, NN, and LN/NN conditions into two QTLs, which were located in the two regions of about 200 and 350 kb with different favorable alleles. The bins 16, 1301, 1465, 1486, 3464, and 6249 harbored the QTLs for NDT detected in this study, and the QTLs/genes previously identified for NDT or nitrogen use efficiency (NUE) could be used for enhancing NDT and NUE by marker-assisted selection (MAS).