13174

1125 دكتري

دكتري

اصلاح نژاد دام

اصفهان: دانشگاه صنعتي اصفهان، دانشكده كشاورزي

۱۳۹۶

يازده، [۱۲۶]ص.:‌ مصور، جدول، نمودار

سعيد انصاري مهياري، مهدي سرگلزايي

محمدعلي ادريس، رستم عبداللهي

عباس پاكدل، قدرت اله سعيدي، علي اسمعيلي زاده كشكوئيه

1396/11/14

كتابنامه

كشاورزي

مهندسي كشاورزي

ID1125 دكتري

Improvement of Genomic Evaluation Using Females Genotypic Data in Genomic Breeding Programs of Dairy Cows Elena Dehnavi e dehnavi@ag iut ac ir January 10 2018 Department of Animal Science Isfahan University of Technology Isfahan 84156 83111 IranSaeid Ansari Mahyari s ansari@cc iut ac irMehdi Sargolzaei msargol@uoguelph caMohammad Ali Edriss edriss@cc iut ac irRostam Abdollahi Arpanahi r abdollahi@ut ac irCollege Graduate Program Coordinator Mehran Shirvani shirvani@cc iut ac ir Department of Animal Science Isfahan University of Technology Isfahan 84156 83111 Iran Centre for Genetic Improvement of Livestock University of Guelph Guelph ON Canada Semex Alliance Guelph ON Canada Department of Animal and Poultry Science Abureyhan University of Tehran Tehran IranAbstract This study was aimed to assess the impact of inclusion of cow data in the trainingpopulation under scenarios encompassing different selection strategies number ofgenotyped cows existence of preferential treatment for elite cows and heritability onaccuracy and bias of genomic predicted values It was also aimed to compare theimputation accuracy of un genotyped sires using the relatives and to assess the accuracyof genomic prediction using imputed sires First a four pathway dairy cattle populationwas simulated for two traits with a low 0 05 and moderate 0 3 heritability Differentnumbers of cows 2500 to 20000 were selected and used in the training population Inthe first study cows were selected randomly R based on the highest H the mostaccurate and highest AH extremes TT or the most accurate from extremes ATT ofbreeding values distribution The simulation was repeated 20 times for each scenario Preferential treatment PT was simulated by introducing upward bias equal to 35 ofphenotypic variance to 5 10 and 20 of elite cows Different sources of informationincluding progeny 4 to 40 progeny per sire both parents and mates were used forimputing ungenotyped sires The ATT scenario resulted in a higher accuracy of GEBV lower bias compared to other scenarios R and TT scenarios followed ATT Usingsimulated data this study showed an improvement in genomic predictions by adding cowinformation into the training population in terms of accuracy and bias Preferentialtreatment of elite cows deteriorated both prediction accuracy and bias Random selectionof cows which might include preferentially treated cows did not reduce the bias Genotype information of cows as mates beside little information of relatives could beinformative to impute ungenotyped sires and increase genomic prediction Keywords Cows selective genotyping genomic evaluation imputation preferential treatment selected validation population Introduction Genomic selection has become a common practice in dairy cattle breeding programsdue to an ever reducing cost of high throughput genotyping products 15 In somecountries genomic evaluation is mainly based on a bull training population because ofthe higher accuracy of bull proofs One problem with a bull only training population isselection bias due to intense selection on bulls 8 and 13 In addition countries with

سعيد انصاري مهياري، مهدي سرگلزايي

محمدعلي ادريس، رستم عبداللهي

عباس پاكدل، قدرت اله سعيدي، علي اسمعيلي زاده كشكوئيه