發(fā)布時間：2018-01-22 05:55

  本文關(guān)鍵詞： 西門塔爾牛 低密度標記 基因組預(yù)測 交叉驗證　出處：《吉林農(nóng)業(yè)大學》2015年碩士論文　論文類型：學位論文

【摘要】：基因組選擇作為動物遺傳育種領(lǐng)域的研究熱點,在動物育種工作中得到了廣泛的應(yīng)用,并對肉牛育種工作產(chǎn)生了巨大的推動作用。但考慮到成本問題,利用高密度SNP芯片進行基因組選擇在實際育種中并不適用。因此,研究者提出采用低密度芯片進行基因組選擇,以降低育種成本。本研究以1,059頭出生于2008至2012年的西門塔爾牛為研究群體,利用Illumina770k SNP芯片,針對宰前活重(Body weight,BW)、胴體重(Carcass weight,CW)和育肥期日增重(Average daily gain,ADG)三個性狀,開展了低密度芯片基因組選擇的研究。研究中利用均勻抽取、基于BayesB所得效應(yīng)值大小和基于GWAS所得P值大小三種不同篩選方式,分別篩選出三種不同類別的低密度芯片(均勻抽取的ELD芯片、根據(jù)效應(yīng)值大小篩選的SLD芯片和根據(jù)P值大小篩選的PLD芯片),進行低密度芯片的基因組預(yù)測研究,并采用交叉驗證方法評價預(yù)測準確性。結(jié)果表明,隨著標記數(shù)目的增多,三種類別的低密度芯片基因組預(yù)測準確性均呈現(xiàn)逐漸增加趨勢。且基于BayesB估計效應(yīng)值大小值篩選的SLD低密度芯片效果要好于另兩種芯片。當SNP標記數(shù)目達到10,000時,在BayesB方法下,SLD低密度芯片三個性狀的基因組預(yù)測準確性分別達到了宰前活重0.22±0.01,胴體重0.21±0.02和平均日增重0.15±0.01。在不同性狀中,三種低密度芯片所表現(xiàn)出的預(yù)測能力不同,宰前活重和育肥期平均日增重兩個性狀中,SLD芯片的預(yù)測準確性,除GBLUP方法下較低外,其他均高于另兩種芯片。這就說明,不同類型低密度芯片基因組預(yù)測能力與目標性狀的遺傳結(jié)構(gòu)有關(guān),因此針對不同應(yīng)用情況仍需進行詳細的育種學分析。本研究以西門塔爾牛為群體,針對宰前活重、胴體重和育肥期增重三個性狀,設(shè)計了不同類型低密度芯片,并對不同低密度標記基因組選擇進行了系統(tǒng)的研究,探討了相關(guān)問題,為制作準確度高、使用方便的低密度芯片和肉牛低密度芯片基因組選擇的實施提供了依據(jù)。
[Abstract]:As a research hotspot in animal genetics and breeding, genome selection has been widely used in animal breeding, and has played an important role in beef cattle breeding. Using high-density SNP microarray for genome selection is not suitable in actual breeding. Therefore, the researchers proposed to use low-density microarray to select genome to reduce the breeding cost. A total of 0 59 Simmental cattle, born between 2008 and 2012, were used to study body weight using Illumina770k SNP chip. BWN, Carcass weight (CW) and average daily gain (ADG) in fattening stage. The genome selection of low density microarray was studied. There were three different screening methods: uniform extraction, effect value based on BayesB and P value based on GWAS. Three different types of low-density chips were selected (evenly extracted ELD chips, SLD chips selected according to the size of the effect value and PLD chips screened according to the P value). The genome prediction of low density microarray was studied and the accuracy of prediction was evaluated by cross validation. The results showed that the number of markers increased with the increase of the number of markers. The accuracy of genome prediction of the three kinds of low density chips showed an increasing trend, and the SLD low density chip based on the size of the estimated effect value of BayesB was better than the other two chips. When SN was used, the accuracy of the low density chip was better than that of the other two kinds of chips. The number of P markers was 10. At #number0#, the accuracy of genome prediction for the three traits of low density microarray was 0.22 鹵0.01 under BayesB. The carcass weight was 0.21 鹵0.02 and the average daily gain was 0.15 鹵0.01. Among different traits, the three low-density microarrays showed different predictive abilities. The prediction accuracy of SLD-chip was higher than that of the other two microarrays except for the lower one under GBLUP method in the two traits of live weight before slaughter and average daily gain during fattening period. Different types of low-density microarray genome prediction ability is related to the genetic structure of target traits, so it is still necessary to carry out detailed breeding analysis for different applications. In this study, Simmental cattle were selected as the population. Different types of low density microarray were designed for three traits, I. E. live weight, carcass weight and weight gain during fattening period, and the selection of different low density marker genomes was studied systematically, and the related problems were discussed. It provides a basis for the implementation of genome selection of low density chips and beef cattle low density chips with high accuracy and convenient use.
【學位授予單位】：吉林農(nóng)業(yè)大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：S823
，

本文編號：1453878