| Dissecting genetic basis of quantitative traits plays an important role in the breeding of crops. However, some shortcomings for the DOS version of segregation analysis software of quantitative traits, i.e., operability, algorithms, functional modules and exist. To overcome these shortcomings, it is necessary to develop the user-friendly Windows interface software under the framework of the Microsoft Visual Studio2010platform and C++programming language.The new version of segregation analysis software for quantitative traits is called SEA. This is because it is derived from the first two letters of segregation and the the first letter of analysis. The main contents in the SEA are as follows.1) As compared with the DOS version of the above software, the new one has been improved in the below aspects, adding model selection function; using the Clapack V3.1.1to solve linear equations and to obtain least squares estimates for genetic parameters; calling Boost V1.51.0library to calculate the likelihood function, probability and homogeneity testing probability; precisely calculating the probabilities in the the Smirnov test(nW2) and Kolmogorov test (Dn); outputting all the results into an Excel file; and calculating posterior probabilities of major-gene genotypes for each individual or family. Using the above algorithms, all the results are consistent with those from SAS software. The SEA is more stable and easier to operate, its results are more complete, and its interface is much more user-friendly.2) The SEA includes one single generation segregation analysis, joint multi-generation analysis and posterior probability calculation. Among these analyses, the one single generation includes F2(SEA-F2), F2:3(SEA-F3), B1:2and B2:2(SEA-BCF), BIL (SEA-BIL); multiple generations have P1, P2, F1and F2(SEA-G4F2); P1, P2, F1and F2:3(SEA-G4F3); P,, P2, F1, B1:2and B2:2(SEA-G5BCF); and P1, P2, F1, F2and F2:3(SEA-G5).3) Results from Monte Carlo simulation studies showed that relatively high accuracy and power for the above segregation analysis softwares are observed, although joint multi-generation analysis has relatively better than one single generation analysis. In addition, the estimates for component variance, dominant effect and major-gene heritability have slightly worse accuracies than those for others, although the formers are acceptable. |
PDF Full Text Request