|WEN, JIA - Nanjing Agricultural University|
|BU, SU-HONG - Nanjing Agricultural University|
|WU, GUORONG - Nanjing Agricultural University|
|XIANG, DAN - Nanjing Agricultural University|
|YI, CAN - Nanjing Agricultural University|
|ZHANG, JIN - Nanjing Agricultural University|
|ZHANG, TIANZHEN - Nanjing Agricultural University|
|ZHANG, YUAN-MING - Nanjing Agricultural University|
Submitted to: Scientific Reports
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/17/2015
Publication Date: 12/17/2015
Citation: Wen, J., Bu, S., Wu, G., Xiang, D., Yi, C., Zhang, J., Song, Q., Zhang, T., Zhang, Y. 2015. Genetic dissection of heterosis using epistatic QTL mapping in partial NCII mating design. Scientific Reports. 5:18376. doi: 10.1038/srep18376.
Interpretive Summary: Heterosis refers to the existence of superior levels of biomass, stature, growth rate and/or fertility in hybrid offspring compared with the parents. The discovery of heterosis in maize a century ago has revolutionized plant and animal breeding and production. Heterosis may be useful for improving soybeans in the future. In this study, we tested computer models that to find quantitative trait loci for heterosis in simulated genomes of self-pollinating crop species. In addition, we proposed statistical methods to reliably estimate their quantitative effects. This study may provide plant breeders at universities, companies and government agencies with new knowledge to utilize crop heterosis for the production of high yield hybrid seed.
Technical Abstract: Heterosis refers to the phenomenon in which hybrid F1 exhibits enhanced growth or agronomic performance. However, theoretical studies on the genetic basis of heterosis were based on bi-parental segregation populations instead of multiple-parental hybrid F1 populations. In simulation study, we mapped quantitative trait locus (QTL) of additive, dominance and epistatic effects in the partial NCII mating design that consists of F1 individuals between two groups of parents using the models with each of the four different dependent variables: trait phenotype, general combining ability (GCA), specific combining ability (SCA) and mid-parental heterosis (MPH). The main and epistatic effects of the QTL in each genetic model were estimated by empirical Bayesian LASSO. The results showed that: 1) GCA consists of additive and additive-by-additive effects, and SCA and MPH consist of dominant and dominant-related effects. The additive-by-additive effect was a small component of SCA; 2) the impact order of components affecting the heterosis was dominant > dominant-by-dominant > over-dominant > complete dominant; and 3) increasing the number of hybrid F1 in mapping population could significantly increase the power to detect dominant-related effects but slightly reduce the power to detect additive and additive-by-additive effects. Analysis of a real dataset showed that more additive QTL were detected in GCA model than in trait phenotype model, two common additive-by-additive QTL were identified in both GCA and trait phenotype models, one dominant-by-dominant QTL was detected in trait phenotype, SCA and MPH models, and more additive-by-dominant QTL were detected in trait phenotype model than in SCA or MPH model. These results provide valuable information for the utilization of crop heterosis.