Prediction Model of Hybrid Performance Using Molecular Marker Based on Additive-Dominant Effects

XU Xin-fu,TANG Zhang-lin,LI Jia-na,CHAI You-rong,WANG Rui,CHEN Li
2008-01-01
Abstract:【Objective】 Mathematic model was explored to predict hybrid performance using molecular markers in Brassica napus L【.Method】 66 hybrids were developed between six recessive nuclear male sterile lines and 11 restorer lines according to the design of North Carolina mating design Ⅱ (NCⅡ). Specific marker loci of SSR and AFLP with significant effects on nine agronomic traits of hybrids were chosen from DNA polymorphic loci among parents by ANOVA to estimate the effects of markers (additive or dominant) and the marker-type values of parents and their hybrids. The relation between the marker-type values and hybrid performance was analyzed. The mathematic models to predict hybrid performance were established by stepwise regression analysis. 【Result】 Marker loci with significant effects of additive or dominant (P=0.01), ranging from 39 to 85, were chosen from 114 SSR and 205 AFLP markers for nine agronomic traits, which exhibited different effects in the size and direction. High and significant (P<0.01) relations of marker-type values were found with hybrid phenotype values with r ranging from 0.6824 to 0.8113 for nine traits. The models established with 6-14 marker loci can be used to predict precisely the hybrid performance, with R2 ranging from 0.5191 to 0.6783 for nine traits.【Conclusion】 It is a new and effective way to predict hybrid performance using marker-type values. It is possible to establish model to predict hybrid performance using small amounts of molecular markers in B. napus L..
What problem does this paper attempt to address?