Representing genetic variation with synthetic DNA standards

The identification of genetic variation with next-generation sequencing is confounded by the complexity of the human genome sequence and biases that arise during library preparation, sequencing and analysis. We have developed a set of synthetic DNA standards, termed ‘sequins’, that emulate human genetic features and constitute qualitative and quantitative spike-in controls for genome sequencing. Reads derived from sequins align exclusively to an artificial in silico reference chromosome, rather than the human reference genome, allowing them to be partitioned for parallel analysis. Here we use this approach to represent common and clinically relevant genetic variation, ranging from single nucleotide variants to large structural rearrangements and copy number variation. We validate the design and performance of sequin standards by comparison to examples in the NA12878 reference genome and demonstrate their utility during the detection and quantification of variants. We provide sequins as a standardized, quantitative resource against which human genetic variation can be measured and diagnostic performance assessed.利用下一代测序技术鉴定遗传变异时，常会受到人类基因组序列的复杂性，以及文库制备、测序与分析过程中产生的系统偏差的干扰。我们开发了一套合成DNA标准品，命名为sequins（sequins），该标准品可模拟人类遗传特征，可作为基因组测序的定性与定量spike-in对照。由sequins产生的测序读段仅能比对至人工计算机模拟参考染色体，而非人类参考基因组，因此可将其分离以进行平行分析。本研究利用该方法覆盖了从单核苷酸变异、大片段结构重排到拷贝数变异等各类常见且具有临床意义的遗传变异。我们通过与NA12878参考基因组中的已知变异实例进行比对，验证了sequins标准品的设计与性能，并证实其在变异检测与定量分析中的应用价值。我们将sequins作为一套标准化的定量资源对外发布，可用于人类遗传变异的定量检测以及诊断性能的评估。