【作者】 马云龙；

【导师】 张勤； Henner Simianer； 丁向东；

【作者基本信息】 中国农业大学 ， 动物遗传育种与繁殖， 2015， 博士

【摘要】 畜禽基因组选择信号检测有助于揭示人工选择作用的潜在遗传机制,进而捕捉造成群体间表型差异的重要候选基因。随着高通量分型技术的发展和成本的降低,一系列基于不同模型的统计方法被开发用于选择信号的检测。这些方法根据使用信息的差异可以分为三大类：基于位点频率谱的方法、基于单倍型的方法与基于群体分化的方法。假如不同的统计方法能够正确检测或者部分正确地检测到基因组存在的选择信号或者由于统计方法之间信息来源和算法设计的相似性,各统计方法之间的统计量或对应的P值应该会存在一定程度的相关。因此,使用多方法策略或进行统计量复合,充分利用不同方法之间的互补,将有助于更加准确、全面地揭示基因组范围内存在的选择信号。从统计学的观点来说,选择一个合适的选择信号检测策略并将不同的选择信号检测方法合并是极具挑战性的。通过大量的数据模拟,本研究比较了目前常用8个选择信号检测方法在不同模拟场景中的表现。研究发现在适当的情况下,间隔为0.1Mb的高密度的分子标记的模拟情景中,选择信号检测方法具有较高的检测效力,而对应的15个个体的小样本也能够获得可靠的检测效力。模拟数据分析显示：CLR、 XPEHH等方法对固定的选择信号具有较高的检测效力,而iHHS方法则对正在进行的选择信号具有较高的检测效力。基于上述结果,本研究构建了一个新的选择信号复合策略——去相关复合统计量(DCMS)。该策略能够校正雇佣方法之间存在可能放大选择信号检测的方法间相关。模拟数据分析显示：DCMS统计量在绝大多数的模拟情景下均具有较高的检测效力和精度。与其它复合方法相比,该方法不仅具有较高的检测效力而且计算过程相对简单。与CMS方法相比,该策略不需要进行复杂的溯祖模拟,并且可以移除统计量之间可能存在的较大且不能忽略的相关。本研究使用DCMS复合策略对人类HapMap的真实数据进行分析,LCT基因的成功捕捉进一步验证了DCMS复合策略的可靠性。基于Illumina猪60K SNP芯片数据,本研究使用四个互补的选择信号检测方法,利用多方法策略对4个猪品种进行全基因组选择信号检测。研究表明：在长白、荣昌、松辽和人白四个品种的基因组上分别发现159、127、179和159个选择信号候选区域,每个区域的平均长度约为0.80Mb、0.73Mb、0.78Mb和0.73Mb。生物信息学分析发现与选择信号候选区域重叠的许多基因或数量性状基因座与繁殖、毛色和耳朵形状等表型性状相关。基于Affymetrix鸡600KSNP芯片数据,本研究雇佣8个常用的选择信号检测方法,使用DCMS复合策略对两个不同皮肤颜色的鸡群体的进行群基因组选择信号检测。研究发现：许多与色素沉积相关的著名基因都有可能与本研究中两个鸡群体的皮肤颜色差异相关联,如BCO2、 MC1R、 ASIP和TYR等基因。选择信号分析作为一个新的可信的畜禽群体基因组学研究手段,全面正确地揭示畜禽基因组范围内存在的选择信号有助于揭示人工选择作用的遗传机理,并且有助于我们更好地制定育种计划。更多还原

【Abstract】 Identifying signatures of selection can provide a straightforward insight into the mechanism of artificial selection and further uncover the causal genes related to the phenotypic variation. With the advent of high throughput and cost-effective genotyping techniques, a series of statistical tests have been developed to detect directional selection signatures based on different models. These methods can be grouped into three categories:site-frequecy spectrum based methods, haplotyped based methods and population differentiation based methods. Viewed from this perspective, statistics and P-values obtained with those methods should exhibit a certain degree of correlation if they reflect fully or partly the same underlying pattern caused by selection or if they are derived from the same basic statistics. Accordingly, it is becoming promising to use multiple methods or a compsite strategy to detect selection signatures to benefit from advantageous complementarities across methods.From a statistical perspective, determining a proper testing procedure and combining various test statistics is challenging. In this study, we discussed the statistical properties of eight different elementary selection signature statistics based on extensive simulations. In the considered scenario we show that a reasonable power to detect selection signatures is achieved with high marker density (>1SNP/kb) as obtained from sequencing, while rather small sample sizes (-15diploid individuals) appear to be sufficient. Most selection signature statistics such as CLR and XPEHH have the highest power when fixation of the selected allele is reached, while iHS has the highest power when selection is ongoing.Furthermore, we suggest a novel strategy, called de-correlated composite of multiple signals (DCMS) to combine different statistics for detecting selection signatures while accounting for the correlation between the different selection signature statistics. When examined with simulated data, DCMS consistently has a higher power than most of the single statistics and shows a reliable positional resolution. Compared to other combining strategies it has the advantage to be easily computable even in populations with not sufficiently known demography (compared to CMS), and to account for correlations of the elementary test statistics, which were found to be too large to be ignored. We illustrate the new statistic to the established selective sweep around the lactase gene in human HapMap data providing further evidence of the reliability of this new statistic.Based on Illumina Porcine60K SNP chip data, four complementary methods were implemented in this study to detect the selection signatures in the whole genome of four pig breeds. In this part, a total of159,127,179and159candidate selection regions with average length of0.80Mb,0.73Mb,0.78Mb and0.73Mb were identified in Landrace, Rongchang, Songliao and Yorkshire, respectively. Bioinformatics analysis showed that the genes/QTLs relevant to fertility, coat color, and ear morphology were found in those candidate selection regions and this analysis also demonstrated the diversity of breeds. Based on Affymetrix chicken600K SNP chip data, we employed eight different elementary selection signature statistics and applied DCMS strategy to scan selection signatures in two chicken samples with diverse skin color. Our analysis suggests that a set of well-known genes such as BCO2, MC1R, ASIP and TYR were involved in the divergent selection for this trait.Selection signature analysis is a relatively novel and highly promising approach in livestock population genomics, an accurate and comprehensive set of selection signatures will be the basis for a better understanding of the forces driving artificial selection and will help to design more efficient livestock breeding programs.更多还原