| Most complex traits of animals,plants are quantitative traits and are associated with human economy.Studies on genetic dissection of quantitative traits are of great importance to the genetic improvement of animals and plants,and to the prevention and control of human complex diseases.Currently,genome-wide association studies is the most commonly used method in studying quantitative traits.With the rapid development of biological sequencing technology,the analysis of millions of markers will become routine.GW AS entails examining a large number of SNPs in a limited sample with hundreds of individuals,implying a variable selection problem in high dimensional dataset.Although many single-locus methods can solve this issue,these methods only detect a small number of QTN due to Bonferroni correction.Also,the total heritabilities of the detected QTN is smaller than the heritability of the traits,implying that a lot of QTN are lost.To address this issue,ISIS EM-BLASSO(Iterative modified-Sure Independence Screening EM-Bayesian LASSO)is one of a series of new algorithms proposed in our laboratory for multi-locus genome-wide association analysis.In order to make the application workers more convenient to use the new method,this research developed the windows interface ISIS EM-BLASSO software package based on R software.The main research contents are as follow:1)The ISIS EM-BLASSO core algorithm and the graphic user interface are organically integrated to form the ISIS EM-BLASSO software.This software is based on the interface design of the data input and the result output.The software depends on additional R packages:gWidgets,gWidgets,gWidgets,RGtk2,RGtk2Extras,cairoDevice,ggplot2,lars and ncvreg packages.gWidgets package provides a convenient and easy platform to create GUI functions.RGtk2Extras package is used because it can display massive amounts of data.RGtk2 package provides a dynamic display scroll bar function.cairoDevice package is embedded in the custom interface R graphics.The core code of multi-locus ISIS EM-BLASSO algorithm relies on ncvreg and lars R packages for its implementation.To draw the LOD score plot,the graphical toolkit GTK + and the drawing package ggplot2 are adopted.The user can change the plot size,plot accuracy,plot resolution rate and LOD line color of the detected significant QTN.The developed ISIS EM-BLASSO package is uploaded to the https://cran.r-project.org/web/packages/mrMLM/index.html website.The R official website converts the R code into Mac,Windows and Linux mainstream operating systems so that ISIS EM-BLASSO software package can run on these operating systems.The ISIS EM-BLASSO software package does not need the help other computer languages,to ensure the simplicity of the software,standardization,and stability.2)Using the ISIS EM-BLASSO and the pLARmEB method,three traits related to flowering time in Arabidopsis thaliana were analyzed.These traits include the flowering time at 10℃,16℃ and 22℃ respectively(FT10,FT16,FT22).The number of significant SNPs detected by ISIS EM-BLASSO method was 79 which is 26 more SNPs than those detected by the pLARmEB method.The total number of genes identified in the proximity of these significant SNPs was 35 which is 19 more genes than those found when pLARmEB method was used.Using the ISIS EM-BLASSO method and the FASTmrEMMA method,two other Arabidopsis traits associated with flowering time were analyzed,that is flowering time with 8 weeks growth in greenhouse(8WGHFT)and leaf number at flowering time with 8 weeks growth in greenhouse(8WGHLN).49 significant SNPs were detected using the ISIS EM-BLASSO method.The ISIS EM-BLASSO method detected 18 more SNPs than the FASTmrEMMA method.23 genes were reported in the proximity of these significant SNPs when ISIS EM-BLASSO method was used.The ISIS EM-BLASSO method identified 11 more genes related to the traits of interest than the pLARmEB method.These results show that the ISIS EM-BLASSO method detected many genes as compared with the other method.This verifies the effectiveness of the ISIS EM-BLASSO method.3)By clicking the software Input Datasets button,a dialog box of Input Datasets pops up and genotypic data and phenotypic data of the corresponding format are imported according to the requirements.The format of genotype data includes mrMLM numerical type,mrMLM character type,and Hapmap(Tassel)type.When genotypic and phenotypic data are imported successfully,clicking the Do button formats the datasets or matches the individuals in phenotypic data to the individuals in genotypic data.By clicking the Population Structure button,a dialog box of whether to import the Population Structure appears.If the datasets to be analyzed requires population structure to be included in the association analysis,the user should import the population structure file.When the genotype,phenotype and population structure are successfully imported into the software,the user should set the key P-value(default P-value is 0.01)and then click the Run button.The program will start running and the progress of the software can be run through the progress bar.After the successful running of the program,the results of the association analysis will be obtained.The software enables the user to visualize the results and obtain the LOD score plot.The software is simple to operate as a design concept,which greatly facilitates the use of genetic breeding workers. |
