回交自交系群体4对主基因加多基因混合遗传模型分离分析方法的建作物学报ACTAAGRONOMICASINICA2021,39(2):198206/zwxb/ISSN0496-3490;CODENTSHPA9E-mail:xbzw@DOI:10.3724/SP.J.1006.2021.00198回交自交系(BIL)群体4对主基因加多基因混合遗传模型分别分析方法的建立王金社赵团结盖钧镒*南京农业高校大豆讨论所/国家大豆改良中心/农业部大豆生物学与遗传育种重点试验室(综合)/作物遗传与种质创新国家重点试验室,江苏南京210095摘要:主基因加多基因混合遗传模型是用于分析数量性状表型数据的统计分析方法,该方法便于育种工利用杂种分别世代的数据对育种性状的遗传组成初步推断,制定相应的育种策略,也可用于校验QTL定位所揭示的数量性状的性状遗传组成。回交自交系(BIL)群体是永久性群体,可以进行有重复的比较试验,适用于受环境影响较大的简单性状的遗传讨论。本讨论以BIL群体为对象构建了４对主基因、主基因加多基因分别分析方法的遗传模型,包括2类11个遗传模型。利用基于IECM(iterativeexpectationconditionalmaximization)算法的极大似然分析方法估算各个混合遗传模型中的分布参数,用AIC值和一组适合性测验结果选取最优模型,并从入选模型的分布参数通过最小二乘法估量遗传参数。由1个模拟的随机区组试验对模型进行验证,模拟群体中遗传参数的估量值与设定值之间具有很好的全都性。利用本文建立的模型重新分析大豆回交自交系群体(EssexZDD2315)及其亲本对胞囊线虫(Hetero-deraglycinesIchinohe)1号生理小种的抗性数据后发觉4对主基因模型优于原报道的3对主基因模型,说明本方法的有效性和正确性。关键词:回交自交家系群体(BIL);主基因加多基因混合遗传;分别分析EstablishmentofSegregationAnalysisofMixedInheritanceModelwithFourMajorGenesPlusPolygenesinBackcrossInbredLines(BIL)PopulationsWANGJin-She,ZHAOTuan-Jie,andGAIJun-Yi*SoybeanResearchInstituteofNanjingAgriculturalUniversity/NationalCenterforSoybeanImprovement/KeyLaboratoryforBiologyandGeneticImprovementofSoybean(General),MinisterofAgriculture/NationalKeyLaboratoryforCropGeneticsandGermplasmEnhancement,Nanjing210095,ChinaAbstract:Thesegregationanalysisofmajorgenespluspolygenesisastatisticalmethodforgeneticanalysisofquantitativetraits.Themethodisparticularlyvaluableforplantbreederstousetheirdataaccumulatedfromsegregationpopulationstoestimatethegeneticsystemoftargettraits,whichisnecessaryfordesigningbreedingstrategiesandalsousefulforvalidatingtheresultsofQTLmapping.Thebackcrossinbredline(BIL)populationisoneofthepermanentpopulations,whichissuitableforgeneticanalysisofcomplextraitsandcanbeusedinreplicatedexperiments.ForBILpopulation,theanalyticalproceduresofthreeandlessmajorgenespluspolygenesmixedinheritancemodelshavebeenestablished.Theobjectiveofthepresentstudywastoestab-lishtheanalyticalproceduresofsegregationanalysisforfourmajorgenespluspolygenesmixedinheritancemodelsinBILpopu-lation.Elevengeneticmodelswithfouradditiveand(or)epistaticmajorgenesincludingthosewithoutandwithpolygeneswereestablished.ThecomponentdistributionparametersweresolvedandestimatedbyusingmaximumlikelihoodmethodbasedonIECM(IterativeExpectationConditionalMaximization)algorithm.Amongthepossiblemodels,thebestonewaschosenaccord-ingtoAkaike’sInformationCriterion(AIC)andasetoftestsforgoodnessoffit.Thenthegeneticparametersoftheoptimalmodelwereestimatedthroughtheleastsquaremethod.Fordemonstrationoftheestablishedprocedures,asimulateddatasetofa本讨论由国家重点基础讨论进展方案(973方案)项目(2021CB1184,2021CB1259,2021CB1093),国家高新技术讨论进展方案(863方案)项目(2021AA10A105,2021AA101106),国家自然科学基金资助项目(31071442,32671266),农业部公益性行业专项(202103060),江苏省优势学科建设工程专项和...