ABCA7 VNTR characterization based on raw current PromethION sequencing data

OBJECTIVESWe recently identified strong association between expanded alleles of an ABCA7 VNTR and Alzheimer's disease (AD). Analysis of this tandem repeat (TR), however, is only possible through low-throughput Southern blotting, which precludes further characterization and implementation in the clinic. We aimed to provide a high-throughput long-read sequencing alternative to assess the length, sequence and methylation state of the ABCA7 VNTR.METHODSWe performed whole genome long-read sequencing on DNA of eleven patients with AD or frontotemporal lobar dementia and healthy control individuals using the recently released PromethION® sequencing platform (Oxford Nanopore Technologies®). We subsequently characterized the ABCA7 VNTR, by developing a new pattern recognition algorithm based on dynamic time warping of raw long-read sequencing data. We validated the results with Southern blotting.RESULTSWith a single sequencing run per individual, we were able to detect all VNTR alleles, which ranged from 300 to more than 10000 bases in length. We obtained length estimates with more than 90% accuracy and high precision (5.6% relative standard deviation). We consistently identified alternative TR sequence motifs, allowing distinction of VNTR alleles with homozygous length. Furthermore, our approach capacitates the detection of nucleotide modifications (e.g. methylation) within the TR.CONCLUSIONSWe provide a new long-read sequencing based method to study the ABCA7 VNTR at an unprecedented resolution, enabling a better understanding of the disease-associated mechanisms. In addition, our approach opens the possibility for whole-genome analysis of (expanded) TRs which will lead to identification of novel disease causing variants and improved diagnostics.

研究目标：我们近期发现ABCA7可变数目串联重复序列（ABCA7 VNTR）的扩增等位基因与阿尔茨海默病（Alzheimer's disease, AD）存在显著关联。然而，该串联重复序列（tandem repeat, TR）的分析目前仅能通过低通量Southern印迹法（Southern blotting）完成，这一方式限制了对其开展进一步表征及临床应用转化。本研究旨在开发一种高通量长读长测序（long-read sequencing）替代方案，用以评估ABCA7 VNTR的长度、序列及甲基化状态。
研究方法：我们使用最新发布的普洛米昂®（PromethION®）测序平台（牛津纳米孔科技公司®，Oxford Nanopore Technologies®），对11名阿尔茨海默病或额颞叶痴呆患者以及健康对照个体的DNA进行全基因组长读长测序。随后，我们基于原始长读长测序数据的动态时间规整算法开发了全新的模式识别算法，以此对ABCA7 VNTR进行表征，并通过Southern印迹法对结果进行验证。
研究结果：通过对每个个体开展单次测序，我们成功检测到了所有VNTR等位基因，其长度范围为300至10000余个碱基对。我们得到的长度估算结果准确率超过90%，且精度较高（相对标准偏差为5.6%）。我们始终能够识别出不同的TR序列基序，从而可区分长度纯合的VNTR等位基因。此外，本方法可实现串联重复序列内核苷酸修饰（如甲基化）的检测。
研究结论：我们开发了一种基于长读长测序的全新方法，能够以前所未有的分辨率对ABCA7 VNTR进行研究，有助于更深入地理解疾病相关的致病机制。此外，本方法为（扩增型）串联重复序列的全基因组分析提供了可能，这将助力新型致病变异体的发现以及诊断方法的优化。