您好,欢迎访问三七文档
基因组时代的动物遗传评估技术遗传评估•评估和比较动物个体在遗传上的优劣•为选择优秀种用个体提供依据•动物育种的中心工作表现型=基因型+环境常规遗传评估技术基因(黑箱)表型环境选择遗传评估亲属的表型特点:利用表型进行遗传评估常规遗传评估技术•BLUP方法是常规遗传评估技术的核心yZyXAZZXZZXXXe''uˆˆ''''12a2y:表型信息A:系谱信息eZuXy常规遗传评估技术•对很多重要经济性状十分有效MilkyieldinUSHolsteins678910111219601970198019902000BirthyearFirstlactationyield(1000kg)GeneticPhenotypic美国荷斯坦奶牛产奶量表型及遗传进展加拿大猪100kg体重日龄遗传进展常规遗传评估技术•局限性–当性状表型难以获取或遗传力低时,遗传评估可靠性不高–不能进行早期遗传评估标记辅助选择(MAS)基因(黑箱)表型数据基因信息遗传评估分子遗传学主效基因/QTL选择特点:利用表型和部分基因的信息进行遗传评估基因辅助选择(GAS)•氟烷基因y=Xg+e连锁不平衡标记辅助选择(LD-MAS)yXMgZuey:表型信息A:系谱信息M:标记信息g:可作为固定和随机效应遗传评估技术-MA-BLUP:标记辅助选择(MAS)•应用现状–实际应用不多–应用效果不显著–主要原因:•已被证实具有显著效应的基因或标记有限–(发现并证实一个有效的基因需要很长的时间和很高的成本)•这些基因或标记仅能解释有限的遗传变异–(10~30%遗传变异)基因组选择(GS)•Meuwissen(2001)等首次提出了基因组选择,但在当时并未引起重视•Schafferr(2006)重新指出GS在奶牛育种中的巨大应用价值•GS成为当期家畜育种最热门的研究领域•优势:–可以捕获所有的遗传变异–无需表型信息即可进行遗传评估,极大地缩短世代间隔和育种成本基因组选择•利用覆盖全基因组的高密度标记(SNP)进行个体遗传评估基因(黑箱)标记信息新的标记(黑箱)估计基因组育种值全基因组SNP芯片YearNo.SNPChipsHuman20083,000,0001000kPoultry20052,800,00060KBovine2007~2,000,00054kPig2008~2,000,00060kSheep2007-50k用于遗传评估的数据用于遗传评估的数据10001112200200121110111121111011110011211000201220022201111202101200211122110021112001111001011011010220011002201101120020110102022212112210201001110001122022122211202112012020100202202000021100011202011221112111022011110000212202000221012020002211220111012100111211102112110020102100022000220100020110000220221102211210112111012222001211212220020002002020201222110022222220022121111210021111200110111011200202220001112011010211121211102022100211201211001111102111211021112200010110111020220022111010201112111101120210210212110110221220012110112110120220110022200210021100011100211021101110002220020221212110002220102002222121221121112002011020200122222211221202121121011001211011020022000200100200011110110012110212121112010101212022101010111110211021122111111212111210110120011111021111011111220121012121101022202021211222120222002121210121210201100111222121101基因组选择•基本步骤1.利用一个参考群体估计每个SNP的效应–参考群体:每个个体都有性状表型记录和所有SNP基因型2.利用SNP效应估计值计算候选群体的个体基因组育种值–候选群体:每个个体都有所有SNP基因型基因组选择参考群体候选群体SNP基因型性状表型获得预测方程SNP基因型选留群体在参考群中估计标记效应gi1niiyXge设计矩阵染色体片断遗传效应估计标记效应在候选群体中计算个体gEBVpiiigXEBV1g全部基因组染色体片断设计矩阵染色体片断效应4×计算基因组育种值(gEBV)基因组育种值1+1-1-1+1+25-1+1-1-1-1+42+1-1-1-1-1-22-1+1-1-1=+38-40SNPeffect-00+20+40-20Chr1Chr2Chr3......ChrnKnownSNPsniiigEBV1ˆgXNumberofSNPsSNPalleleeffect标记效应的估计方法最小二乘法BLUP贝叶斯方法BayesA,B,C其他方法半参数、非参数机器学习、主成份分析最小二乘法对标记效应分布无任何假设1:对每个标记作单点回归分析2:选择效应值最大的m个点放入模型中,同时对其进行估计1nyXge1niiyXge不足:1.难以确定模型选择的阈值2.容易高估标记效应BLUP方法1BLUPmethod(1)egXWfynjjj12)(gjVarIg假设所有染色体片段具有相同的效应方差!gj=chromosomalsegmenteffectyXyWgfIλXXWXX2gHowtochose?BLUP方法2BLUPmethod(2)egXZaWfynjjj1yXyZyWgafIλXXZXWXXZAZZWZXWZ=residualpolygeniceffect2)(aVarAaBLUP方法3BLUPmethod(3)egXWfynjjj12)(jjVarIg每个染色体片段有自己的效应方差!改变染色体片段效应的方法ModelvariantFunctionlinearweight(LW)exponentialweight(E1,USDA)exponentialweight(E2,USDA)quadraticweight(Q1)quadraticweightwithlimits(Q2)122jjjsˆ22jsjˆ2212.1jsjˆ2225.1222ˆjjsisstandardisedeffectestimateofmarkerjjsˆBLUP方法4BLUPmethod(4)eZaWfyyZyWafGZZWZZ)(aVarGaGenomicrelationshipmatrixGenotypePedigree121101011110111211120200101121101111122221121111101101111102011111012011121120011010BayesformulaP(A)isthepriorprobabilityormarginalprobabilityofAP(A|B)istheconditionalprobabilityofA,givenB.ItisalsocalledtheposteriorprobabilitybecauseitisderivedfromordependsuponthespecifiedvalueofB.P(B|A)istheconditionalprobabilityofBgivenA.Itisalsocalledthelikelihood.P(B)isthepriorormarginalprobabilityofB,andactsasanormalizingconstant.Bayes'theoremwithcontinuouspriorandposteriordistributionstogettheposteriorprobabilitydistribution,multiplythepriorprobabilitydistributionbythelikelihoodfunctionandthennormalize.Bayesianinferencereferstotheuseofapriorprobabilityoverhypothesestodeterminethelikelihoodofaparticularhypothesisgivensomeobservedevidence;thatis,thelikelihoodthataparticularhypothesisistruegivensomeobservedevidence(theso-calledposteriorprobabilityofthehypothesis)comesfromacombinationoftheinherentlikelihood(orpriorprobability)ofthehypothesisandthecompatibilityoftheobservedevidencewiththehypothesis(orlikelihoodoftheevidence).Bayesian方法BayesianmethodModelforthedataModelforthemarkervariancePriordistributionof2giBayesian方法BayesABayesian方法BayesBBayesian方法BayesC基因组选择在奶牛育种中应用年青后备公牛分型一些AI(人工授精)种公牛站也开始直接根据年青公牛基因组育种值推销精液小母牛和成年母牛的基因分型基因组选择在奶牛育种中应用参考群(Holstein)USA/Canada5000bullsbornbefore2000Germany4339bullsbornbefore2004France1750bullsNetherlands1500bullsAustralia2000bullsIreland~950bullsNewZealand1450bulls,born1980-2001Nordic4000bulls,Born1986-2002Poland1227bullsbornbefore2003基因组选择在奶牛育种中应用用来估计SNP效应的方法USABLUP3CanadaBLUP1GermanyBLUP2FranceBayesBNetherlandsBLUP2AustraliaBayesBIrelandBLUP1NewZealandBLUP1NordicBayesDPolandBLUP1基因组选择在奶牛育种中应用USA2009.1Canada2009.8Germany2009.8France2009.6Netherlands2009.7Australia2010Ireland2009.2NewZealand2008.8Nor
本文标题:基因组选择
链接地址:https://www.777doc.com/doc-5695069 .html