Finding Markers That Make a Difference: DNA Pooling and SNP-Arrays Identify Population Informative Markers for Genetic Stock Identification

Genetic stock identification (GSI) using molecular markers is an important tool for management of migratory species. Here, we tested a cost-effective alternative to individual genotyping, known as allelotyping, for identification of highly informative SNPs for accurate genetic stock identification. We estimated allele frequencies of 2880 SNPs from DNA pools of 23 Atlantic salmon populations using Illumina SNP-chip. We evaluated the performance of four common strategies (global FST, pairwise FST, Delta and outlier approach) for selection of the most informative set of SNPs and tested their effectiveness for GSI compared to random sets of SNP and microsatellite markers. For the majority of cases, SNPs selected using the outlier approach performed best followed by pairwise FST and Delta methods. Overall, the selection procedure reduced the number of SNPs required for accurate GSI by up to 53% compared with randomly chosen SNPs. However, GSI accuracy was more affected by populations in the ascertainment group rather than the ranking method itself. We demonstrated for the first time the compatibility of different large-scale SNP datasets by compiling the largest population genetic dataset for Atlantic salmon to date. Finally, we showed an excellent performance of our top SNPs on an independent set of populations covering the main European distribution range of Atlantic salmon. Taken together, we demonstrate how combination of DNA pooling and SNP arrays can be applied for conservation and management of salmonids as well as other species.基于分子标记的遗传种群识别（Genetic Stock Identification, GSI）是洄游物种管理的重要工具。本研究针对精准GSI所需的高信息含量单核苷酸多态性（Single Nucleotide Polymorphism, SNP）筛选工作，测试了一种替代个体基因分型的低成本方案——等位基因分型（allelotyping）。我们利用Illumina SNP芯片，对23个大西洋鲑种群的DNA混合样本中的2880个SNP位点进行了等位基因频率估算。我们评估了四种常用策略——全局固定指数（Fixation Index, FST）、成对固定指数（FST）、Delta法以及异常值检测法——用于筛选最优信息SNP位点的表现，并与随机选取的SNP和微卫星（microsatellite）标记组相比，测试了这些策略在GSI中的应用效果。在大多数情况下，通过异常值检测法筛选出的SNP位点表现最优，其次为成对FST法与Delta法。总体而言，与随机选取的SNP位点相比，本筛选流程可将精准GSI所需的SNP位点数量最多减少53%。但GSI的识别精度更多受候选筛选群体的影响，而非排序筛选方法本身。本研究通过整合迄今为止规模最大的大西洋鲑种群遗传数据集，首次验证了不同大规模SNP数据集的兼容性。最后，我们在覆盖大西洋鲑欧洲主要分布范围的独立种群样本集中，验证了筛选得到的最优SNP位点集的优异表现。综上，本研究证明了DNA混合技术与SNP芯片的结合方案，可应用于鲑科鱼类及其他物种的保护与管理工作。