MassIVE.quant: a curated and scalable community resource for quantitative proteomics

项目摘要 该项目将贡献MassIVE.quant，这是一种基于定量质谱的新型数据资源。 蛋白质组学 定量质谱法表征复杂生物混合物中的蛋白质， 准确性、灵敏度和吞吐量。大多数这样的实验的分析涉及肽的鉴定， 产生光谱的蛋白质，以及预定义的蛋白质之间丰度变化的相对定量。 条件虽然鉴定工作流程现在已经成熟，并准备进行可重复的研究， 定量工作流程远远落后。现在，没有任何存储库可以存储跨所有 工作流程，作者通常不可能以允许独立评估的形式提供数据 和再利用。这破坏了这些调查的可重复性和影响。 该项目结合了Banderia实验室在开发质谱互动方面的专业知识 虚拟环境（MassIVE），用于存储、记录和重新分析质谱的公共存储库， 识别，以及Vitek实验室在开发MSstats方面的先前专业知识，MSstats是一个广泛的统计数据集合， 用于定量蛋白质组学工作流程的方法和软件。首先，该项目将全面记录和注释 中等规模的定量调查“培训集”（通常依赖人工程序）， 用于记录和注释关于样品生物来源的实验的标准， 以及数据采集和处理的技术方面。第二，项目将设计功能 对于原始调查的全存储库完整和自动重新分析，使用有限数量的 “良好做法”工作流程。重新分析将完全保留结果的来源，并将用于 进一步描述实验设计和结论中的潜在缺陷。最后，该项目将 将这些研究纳入更广泛的科学背景。它将设计一个查询基础结构， 实验的同行调查，即具有类似生物或技术方面的调查，以提供 对结果一致性的洞察。 继续PI先前广泛的外联工作，将向广大社区传播结果 利益相关者，包括蛋白质组学科学家，工具开发人员，期刊编辑，学员和感兴趣的科学家 蛋白质水平的信息。 该项目将通过创建一个公共资源， 目前还不存在任何形式。它将扩大该领域的技术能力，最终使我们能够 更准确地描述生物学功能。

PROJECT SUMMARY The project will contribute MassIVE.quant, a novel data resource for quantitative mass spectrometry-based proteomics. Quantitative mass spectrometry characterizes proteins in complex biological mixtures with the highest available accuracy, sensitivity and throughput. Analysis of most such experiments involves identification of peptides and proteins that generated the spectra, and relative quantification of changes in abundance between pre-defined conditions. While the identifications workflows are now mature and ready for reproducible research, the quantitative workflows lag very far behind. No repositories can now store the analyses results across all workflows, and it is often impossible for authors to provide their data in a form that allows independent evaluation and reuse. This undermines the reproducibility and the impact of these investigations. The project combines the prior expertise of the Banderia’s lab in developing Mass spectrometry Interactive Virtual Environment (MassIVE), a public repository for storing, documenting and re-analyzing mass spectra for identification, and the prior expertise of the Vitek lab in developing MSstats, a broad-scope collection of statistical methods and software for quantitative proteomic workflows. First, the project will fully document and annotate a medium scale “training set” of quantitative investigations (which often rely on manual procedures), to develop standards for documenting and annotating the experiments with respect to the biological origins of the samples, and the technological aspects of data acquisition and processing. Second, the project will design functionalities for repository-wide complete and automated re-analyses of the original investigations, using a limited number of “good practice” workflows. The re-analyses will fully preserve the provenance of the results, and will be used to further characterize potential pitfalls in the experimental designs and conclusions. Finally, the project will place these investigations into a broader scientific context. It will design a query infrastructure that links each experiment to its peer investigations, i.e. investigations with similar biological or technological aspects, to provide insights into consistency of the results. Continuing the extensive prior outreach efforts of the PIs, the results will be disseminated to a broad community of stakeholders, including proteomic scientists, tool developers, journal editors, trainees, and scientists interested in protein-level information. The project will shift the mass spectrometry-based research paradigm, by creating a public resource that currently does not exist in any form. It will expand the technical capabilities of the field, ultimately allowing us to make more accurate of statements about the biological function.