Abstract: Studies on inheritance of quantitative traits are usually based on two-way crosses, while in plant breeding, knowledge on inheritance of traits among a group of parental materials is needed since elite genes or QTLs are scattered in different materials. It is especially true in breeding for hybrids. Based on our previous studies on major locus group analysis among a set of parents, we aimed at extending the analysis to major-minor locus groups in diallel cross experiment. The statistical method was derived from a forward stepwise regression for identifying major locus groups and estimating their genetic effects as well as the estimation of genetic effect of collective minor locus groups. In which the procedure of adjusting F criterion was raised since biased F-test was found in a simulation study. Then the procedure was used to evaluate the relationship between its power and some factors, including heritability value, number of major locus groups, genotypic frequency with increment alleles, similarity among major locus groups and effect of collective minor locus groups. A simulated diallel cross experiment was used to demonstrate the effectiveness of the procedure. A diallel cross experiment on days to flowering of rapeseed was given to show the statistical steps and its utilization. The procedure of major-minor locus group analysis in diallel crosses based on major gene plus polygene inheritance model provides a way to analyze the genetic constitution of a group of parental materials for obtaining the best hybrid combinations.

Key words: Diallel cross design, Major gene and polygene model, Major locus group, Minor locus group, Stepwise regression