Development Of Windows Version Of Segregation Analysis Software And Detection Of Imprinting QTL For Quantitative Traits

Phenotypic observations in bi-parental segregation populations play an important role in dissecting genetic basis of quantitative traits and crop breeding.Our previous version of segregation analysis software of quantitative traits has been used for several decades in domestic and overseas. The application research reveals some shortcomings in its operability, algorithms and functional modules. Therefore, it is necessary to develop the fully functional and user-friendly segregation analysis software for Windows interface. The new software (SEA) has been improved in interface design, function module and algorithm. In terms of interface design, with Windows interface and interaction functions, it is simple and easy to import data, set parameters, compute and save results. In terms of function module, some improvements are as following:(1) Adding model select function; (2) Adding the first-order and second-order genetic parameter calculation; (3) Calculating posterior probabilities of major-gene genotypes for each individual or family; (4) Regulating the output format and outputting all the results into an Excel file. In terms of algorithms, using the Clapack V3.1.1 to solve linear equations and obtain the least squares estimates for genetic parameters; calling Boost V1.51.0 library to calculate the likelihood function, probability and homogeneity testing probability; precisely calculating the probabilities in the Smirnov test (nW2) and Kolmogorov test (Dn).The new version software includes one single generation segregation analysis, joint multi-generation analysis and posterior probability calculation for segregating generations. Among these analyses, the one single generation includes DH (SEA-DH), B1 and B2 (SEA-BC); multiple generations have P1, P2 and DH (SEA-G3DH); P1, P2, F1 B1 and B2 (SEA-G5BC); P,, P2, F1, B1, B2 and F2 (SEA-G6); P1, P2, F1, Bi:2, B2:2 and F2:3 (SEA-G6F). Results from Monte Carlo simulation studies validated the feasibility of the new software.Expression of imprinted genes (iQTL) has an attribute of parental selectivity and is often studies through biological way. In this paper, we use the statistical method to identify such genes in order to provide foundation for biology study. Based on the interval mapping method of iQTL in immortalized F2 population, we developed an empirical Bayes method for mapping multiple QTL. Results from Monte Carlo simulation studies indicated the new method has higher power and accuracy.