多性状全基因组关联分析新方法及其在设计育种中的应用研究

Most crop breeding target traits, such as yield，quality and stress resistance, are quantitative traits, controlled corporately by a set of interacted gene network and environment. Intricate genetic correlation is evolved in phenotypic correlation between traits, especially between comprehensive character and its components. The essential task of crop breeding practice is to comprehensively improve target traits and breed new environmental friendly cultivars, its success or not depends on dissection of genetic architecture and identification of functional genes of traits. With development of high throughput technology, the linkage analysis methods are changing to association methods for QTL mapping. Currently, the available association methods are for separate analysis of each trait, couldn't analyze the gene-gene interaction effects and gene-environment interaction, as well as pleiotrophic effects of gene. In the statistical framework of mixed linear model, our research will develop new association and conditional association method for mapping genes based on integrated analysis of multiple traits, design new method to fast and efficiently screen candidate gene set in genome wide potentially associated with surveyed characters, for dissecting genetic architecture based on multiple omics data of phenotype, gene expression, protein and metabolism, and analyzing each genetic component effects of gene-gene interaction, gene-environment interaction, accordingly, a software will be developed in parallelized computation technology. The expected results of this research are very valuable for dissection of genetic architecture, identification of functional genes and their breeding application in science and practice of genetic improvement of quantitative traits.

作物产量、品质、抗逆等育种目标性状都是数量性状，受多基因互作网络与环境的协同调控，性状间特别是综合性状与构成因素间存在内在遗传组分的相关。综合改良育种目标性状、培育环境友好型新品种是作物育种的根本任务，成功与否取决于性状遗传机理的解析和功能基因的挖掘。随高通量测序技术的发展，数量性状基因定位正从连锁分析向关联分析转变，但当前关联分析方法都针对单一性状，无法全基因组分析基因间、基因与环境间的互作，也无法分析基因的多效性。课题将在混合线性模型框架下，发展多性状关联分析新方法，发展新方法全基因组快速筛选潜在关联的候选SNP集，剖析复杂性状遗传结构，分析基因间、基因与环境互作的各项效应；发展多性状改良的基因型设计理论与方法；研制并行化分析程序。课题研究成果对作物数量性状遗传结构剖析、功能基因挖掘及育种应用具有重要的科学意义和实践应用价值。

作物产量、品质、抗逆等育种目标性状都是数量性状，数量性状遗传结构的剖析，对性状的遗传改良与新新品种培育具有重要意义。随着高通量测序技术的发展，数量性状基因定位正从连锁分析向关联分析转变，但当前关联分析方法都针对单一性状，功效不高，无法全基因组分析基因间、基因与环境间的互作，也无法分析基因的多效性。课题针对已有关联分析方法存在的不足，开展了多性状关联分析新方法、育种应用及分析软件的研究，主要开展了以下几方面内容的研究：1）针对全基因组高密度分子标记数据，发展了半参数的广义多因子维数缩减法与混合线性模型相结合、全基因组分析复杂性状基因间互作、基因与环境间互作的关联分析新方法，研制配套软件；2）在混合线性模型理论框架下，发展了包含环境、基因上位性、基因与环境互作效应的多性状关联分析模型和分析算法，实现多个遗传相关性状的联合关联分析，为性状遗传相关的分子剖析提供方法；3）发展了分歧性状广义多因子维数缩减法，为动植物离散型性状全基因组分析基因间互作、基因与环境间互作提供分析方法及软件；4）以水稻及棉花为实例，开展了重要农艺与经济性状的全基因组关联分析、条件关联分析，鉴别显著关联的遗传位点，并采用生物信息学方法，筛选具有育种潜在应用价值的功能基因，展开性状改良的基因型分子设计与育种展望评价。课题研究已有3篇研究论文在SCI期刊正式刊出，总影响因子10.339，多性状关联分析方法已申请国家发明专利（201610045592.8），另有1篇稿件已投《Journal of Experimental Botany》，1篇投《Scientific Reports》稿件已经返修。课题发展的方法与软件促进了动植物复杂性状遗传结构有效剖析及功能基因挖掘和育种利用，已经在全国推广应用，举办了两届全国数量遗传分析与QTX定位研讨班，起到了良好社会效果。