Inclusive Composite Interval Mapping Of Quantitative Traitloci By Environment Interactions

QTL by environment interaction(QEI) was widely detected in genetic analysis of crops. Studies on QEI contribute to the efficient use of marker-assisted selection in breeding, and better understanding of genetic architecture of important quantitative traits and genotype by environment interactions. Inclusive Composite Interval Mapping(ICIM) has been proposed to map additive, dominant and epistatic QTL for biparental populations in single environment, illustrated good properties and used widely in actual genetic studies. In this study, ICIM was extended to additive, additive-dominant, and epistatic QTL by environment interaction mapping for multi-environmental trials, and its efficiency was illustrated using simulated and experimental populations. This study included the following four aspects.Firstly, statistical algorithms of additive, additive-dominant and epistatic QEI mapping using ICIM were illustrated in DH and F2 populations. Similar to single-environment analysis, stepwise regression was firstly applied in each environment to identify the most significant marker(and marker pairs for epistatic QEI mapping) variables which were then used to adjust the phenotypic values. One-dimensional(or two-dimensional for epistatic QEI mapping) scanning was then conducted on the adjusted phenotypic values across the environments in order to detect QTL with average additive(or/and dominant, or epistatic) and QEI effects. QEI mapping results included QTL positions, three LOD(logarithm of odds scores) values, i.e., LOD, LODA(or LODAA) and LODAE(or LODAAE), three PVE(phenotypic variance explained) values, i.e., PVE, PVEA(PVEAA) and PVEAE(PVEAAE), average additive(or/and dominant, or epistatic) effects and QEI effects, which directly reflected the significance of QTL, average effect and QEI effect.Secondly, genetic models including independent and linked QTL models were designed to study the efficiency of the proposed methods, which considered different QEI levels, linkage phases, heritability levels and positions of QTL in marker intervals. The results showed that ICIM had high detection power, low false discovery rate(FDR), asymptotically unbiased estimates of QTL positons and effects. The power increased and FDR decreased with the increase in PVE or heritability.Thirdly, the comparison between QEI mapping and single-environment mapping in two maize populations demonstrated the properties of QEI mapping. QTL stability and QEI levels can be analyzed directly by referring to QEI mapping results. Most QTL detected by single-environment mapping can also been detected by QEI mapping. Estimates of QTL effects and positions were similar in both mapping methods, but estimates from QEI mapping were more reliable. Only stable QTL could be detected by QTL mapping using average phenotypic performance across environments.Finally, an empirical method to determine the LOD threshold was extended to QEI mapping. The empirical LOD threshold can be determined by formula)(/)( 1022 ln LODplca=, where lc)(a2p is the inverse 2c distribution that returns the critical value of LRT(likelihood ratio test) of a right-tailed probability pa for the degree of freedom l. In QTL mapping, l was equal to the product of the number of genetic parameters to be estimated and the number of environment. Key words: Multi-environmental trials, Inclusive composite interval mapping(ICIM), QTL by...