Improving metagenomic analysis with novel algorithms and technologies

总结/摘要 通过测序进行的微生物组研究对临床研究越来越重要。 人体肠道微生物组已被证明具有各种各样的潜在健康 影响。然而，我们对微生物进行遗传分析的能力仍然有限。微生物组是 非常复杂和标准的短读段测序技术通常不能提供 为恢复相关基因和生物体提供了充分的依据。 低投入、低成本的连接阅读DNA测序技术，如10 x Genomics 铬系统，最近出现了前所未有的承诺从头组装 全基因组或宏基因组样品。这些技术采用了一种新的分子 条码技术，提供了远距离的信息超过标准的高通量短 读取，下一代测序，同时仍然在合理的试剂和低成本。我们计划 开发几种创新的新算法，以快速的方式充分利用条形码读取， 改进几个完整和具有挑战性的应用，特别是：改进宏基因组 组装和利用对低丰度基因组信息的增加的敏感性， 为了识别临床相关的和潜在的致病微生物， 决策 我们提出的所有方法和计算工具都将免费提供， 文档供社区使用。为了确保我们的方法的实用性，我们计划 将它们广泛应用于广泛的研究和临床鸟枪宏基因组数据集， 在我的实验室，并通过各种建立的地方，外部和工业合作。 我们还计划收集对照样品并使用多个平台（Illumina， 10x Genomics，Loop Genomics Read Cloud，UTS TELL-SEQ，Oxford Nanopore） 基准测试。我们还将使用我们提出的方法来提高检测和 临床样品中低丰度生物体的分类。我们将启动两个试点项目 与我们的病理科和特殊外科医院（HSS）合作。 该项目的成功完成将提供快速和可扩展的计算方法， 可以应用于大规模数据集。

SUMMARY/ABSTRACT Microbiome research through sequencing is becoming increasingly important for clinical studies. The human commensal microbiomes have been shown to have a wide variety of potential health impacts. However, our ability to genetically assay microbes is still limited. Microbiomes are extremely complex and standard short-read sequencing technologies often do not provide sufficient basis for the recovery of relevant genes and organisms. Low-input and low-cost linked-read DNA sequencing technologies, such as the 10x Genomics chromium system, have recently emerged with unprecedented promise for de novo assembly of whole genome or metagenome samples. These technologies employ a novel molecular barcoding technique which offers long-range information over standard high-throughput short read, next-generation sequencing, while still at reasonable reagent and low-costs. We plan to develop several innovative novel algorithms to fully leverage barcoded reads in a fast manner to improve several integral and challenging applications, in particular: improving metagenome assembly and leveraging the increased sensitivity to low abundance genomic information in order to identify clinically relevant and potentially pathogenic organisms that can inform clinical decisions. All our proposed methods and computational tools will be made freely available with extensive documentations for the community to use. To ensure the utility of our methods we plan to extensively apply them to a wide range of research and clinical shotgun metagenome data sets, in my laboratory and through various established local, external and industrial collaborations. We also plan to collect control samples and sequence them using multiple platforms (Illumina, 10x Genomics, Loop Genomics Read Cloud, UTS TELL-SEQ, Oxford Nanopore) for benchmarking. We will also use our proposed methods to improve the detection and classification of low abundance organisms in clinical samples. We will launch two pilot projects in collaborations with our Department of Pathology and Hospital for Special Surgery (HSS). Successful completion of this project will provide fast and scalable computational methods that can be applied to large-scale data sets.