Development And Application Of Genomic Prediction Software Package Predhy

Genomic selection(GS)is a new technique to predict unknown phenotypes by using genome-wide markers.With the development of high-throughput sequencing technology,GS has been widely used in crop breeding.The genotype of hybrids can be inferred from the parental genotype in the hybrid breeding of crops,so the advantages of GS are even more prominent.However,there are still many problems in the application of GS in crop breeding practice,such as complex analysis process,difficult model selection and high genotyping cost limit the large-scale breeding application of GS,so the development of convenient,flexible and practical genomic prediction software and the selection of high-quality and affordable genotyping platforms are the current research hotspots and difficulties.In this study,we developed a software package for Genomic Prediction of Hybrid Performance and explored the effects of different genotyping platforms for genomic prediction of maize hybrid performance.1.Development of the genomic prediction software package predhy.This study uses existing R-libraries(BGLR,pls,glmnet,randomForest and xgboost),an integrated R toolkit for GS named predhy(Genomic Prediction of Hybrid Performance)was developed.The software mainly consists of four modules:(1)Genotype data quality control and format conversion;(2)Inference of hybrid genotype;(3)Cross-validation;(4)Prediction of hybrid phenotype.Using two maize data sets as examples,the function of predhy has been demonstrated and evaluated.Due to the complexity of genetic structure,no single method or model is robust for all traits.The predhy mainly selects the most appropriate methods and models through the cross-validation process,predicts the phenotype of untested hybrids,and then selects the excellent crosses,which provides a reference for user to perform GS analysis.2.The effects of different genotyping platforms on genomic prediction of maize hybrid performance.GS is a powerful tool for improving genetic gain in maize breeding.However,its routine application in large-scale breeding pipelines is limited by the high cost of genotyping platforms.Although sequencing-based and array-based genotyping platforms have been used for GS,few studies have compared prediction performance among platforms.In this study,we evaluated the predictabilities of four agronomic traits in 305 maize hybrids derived from 149 parental lines subjected to genotyping by sequencing(GBS),a 40KSNP array,and target sequence capture(TSC)using eight GS models.The GBS marker dataset yielded the highest predictabilities for all traits,followed by TSC and SNP array datasets.We investigated the effect of marker density and statistical models on predictability among genotyping platforms and found that 1KSNPs were sufficient to achieve comparable predictabilities to 10K and all SNPs,and BayesB,GBLUP,and RKHS performed well,while XGBoost performed poorly in most cases.We also selected significant SNP subsets using genome-wide association study(GWAS)analyses in three panels to predict hybrid performance.GWAS facilitated selecting effective SNP subsets for GS and thus reduced genotyping cost,but depended heavily on the GWAS panel.We conclude that there is still room for optimization of the existing SNP array,and using genotyping by target sequencing(GETS)techniques to integrate a few functional markers identified by GWAS into the 1KSNP array holds great promise of being an effective strategy for developing desirable GS breeding arrays.