Analyzing Next-Generation Sequencing Data

现代生物医学研究越来越多地利用来自下一代测序平台的基因组规模数据，包括Illumina HiSeq和MiSeq机器以及Pacific Biosciences SMRT。这些平台使各个实验室能够快速、廉价地从从头测序、重测序、ChIP-seq、mRNA-seq和等位基因分型实验中生成大量的基因组和转录组数据。尽管生物学家有能力快速生成大型数据集，但他们很少接受过必要的计算和统计技术培训，以理解这些数据。因此，许多研究人员必须依靠其他人-通常是几乎没有生物学训练的计算科学家-来设计和实施适当的数据简化和数据挖掘技术。此外，大多数机构无法获得运行这些分析所需的大量计算资源。我们将继续通过为期两周的短期强化夏季课程来帮助弥合这一差距，方法是教授生物医学研究人员（1）在托管在Amazon Web Services“云”上的远程UNIX服务器上运行分析;（2）对大型短读数据集执行映射和组装;（3）利用现有短读数据解决特定的生物学问题;以及（4）设计能够解决他们自己的研究问题的计算管道。这将通过相关技术的深入实践培训来实现。我们的经验，通过评估证实，是这种实际的培训导致参与者的基本计算复杂性的实质性改善。我们相信，从长远来看，我们的干部和其他课程将有助于在数据驱动生物学的一般领域的显着改善。这个短期课程将继续帮助培训当前和下一代独立的生物医学研究人员在基本的计算思维和程序，以及教他们如何利用可扩展的互联网计算资源为自己的研究。此外，我们将继续开发和扩展我们广泛的在线材料，这些材料可以在网上免费获得并广泛使用。我们的最终目标是提高生物医学研究人员利用新型测序技术的效率和复杂性。对于这次更新，我们建议继续以低成本提供课程;扩大我们的RNAseq讨论;通过扩大学习编程和UNIX的可用材料来满足学生的需求;并显着增加我们的统计部分。

DESCRIPTION (provided by applicant): Modern biomedical research is increasingly making use of genome-scale data from next-generation sequencing platforms, including Illumina HiSeq and MiSeq machines and Pacific Biosciences SMRT. These platforms make it possible for individual labs to quickly and cheaply generate vast amounts of genomic and transcriptomic data from de novo sequencing, resequencing, ChIP-seq, mRNA-seq, and allelotyping experiments. Despite this ability to quickly generate large data sets, biologists are rarely traine in the computational and statistical techniques necessary to make sense of this data. Thus, many researchers must rely on others - often computational scientists with little biological training - to design and implement appropriate data reduction and data mining techniques. Moreover, most institutions do not have access to the substantial computational resources necessary to run these analyses. We will continue to help bridge this gap with a short, two-week intensive summer course, by teaching biomedical researchers to (1) run analyses on remote UNIX servers hosted in the Amazon Web Services "cloud"; (2) perform mapping and assembly on large short-read data sets; (3) tackle specific biological problems with existing short-read data; and (4) design computational pipelines capable of addressing their own research questions. This will be accomplished by in-depth hands- on practical training in the relevant techniques. Our experience, confirmed by assessment, is that this practical training leads to a substantial improvement in the basic computational sophistication of participants. We believe that in the long term our cadre and those of other courses will contribute to a significant improvement in the general area of data-driven biology. This short course will continue to help train the current and next generation of independent biomedical researchers in basic computational thinking and procedure, as well as teaching them how to make use of scalable Internet computing resources for their own research. Moreover, we will continue to develop and extend our extensive online materials, which are freely available online and widely used. Our end goal is increase the efficiency and sophistication with which biomedical researchers make use of novel sequencing technologies. For this renewal, we propose to continue offering the course at a low cost; expand our RNAseq discussion; address student needs by expanding the available materials for learning programming and UNIX; and increase our statistics component significantly.