Computational tools for mass spectrometry-based interactome data

描述(申请人提供)：对蛋白质复合体和相互作用网络及其动态行为的分析在生物学研究中具有重要意义。亲和纯化-质谱联用(AP-MS)是目前广泛用于蛋白质相互作用分析的方法。我们的工作解决了为这些数据开发强大的计算方法和工具的迫切需要。我们已经证明了可以从AP-MS数据中提取的无标记定量蛋白质丰度信息的巨大用途，并开发了AP-MS研究中用于评估蛋白质相互作用的统计分析(SAINT)框架。我们还发起了一个国际联盟，对AP-MS实验中观察到的非特异性结合蛋白进行全面分类-亲和纯化污染物储存库(CRAPome.org)。在这些进展的基础上，我们将继续朝着我们的目标发展，即开发一个全面的计算资源，用于对适用于最常用的实验方案和MS平台的蛋白质相互作用数据进行评分。我们还将更好地了解非特定结合生成知识的潜在机制，这些知识对回顾分析以前发表的数据和设计未来的实验都很有用。通过将实验AP-MS数据与外部信息(如基于结构的蛋白质相互作用预测)相结合，我们将进一步提高Ap-MS的灵敏度 检测低丰度和瞬变相互作用。同样显而易见的是，绘制像人类这样的有机体的完整相互作用图是一项社区范围的努力，多个群体贡献了整个互动组的不同部分。我们将开发一个新的计算框架，以一致地整合来自不同研究的AP-MS数据集，从而产生更完整和准确的定量相互作用网络。最后，一个尚未完全解决的重要问题是定量分析蛋白质复合体和相互作用网络动力学。高度灵敏的多重MS技术的出现提供了这样一个机会，我们将开发先进的计算算法和工具，利用多重MS数据进行差示和动态交互作用组分析。我们将继续向生物界提供我们广泛使用的计算工具和数据资源，以及基准数据集，以供其他科学家进一步开发计算方法。

DESCRIPTION (provided by applicant): The analysis of protein complexes and interaction networks, and their dynamic behavior are of central importance in biological research. Affinity purification coupled with mass spectrometry (AP-MS) is now widely used for protein interaction analysis. Our work addresses the critical need to develop robust computational methods and tools for these data. We have demonstrated the great utility of label-free quantitative protein abundance information that can be extracted from AP-MS data, and developed the Statistical Analysis of INTeractomes (SAINT) framework for scoring protein interactions in AP-MS studies. We have also initiated an international consortium to comprehensively catalogue the non-specific binding proteins observed in AP-MS experiments - the Contaminant Repository for Affinity Purification (CRAPome.org). Building upon these advances, we will continue toward our goal of developing a comprehensive computational resource for scoring protein interaction data applicable to most commonly used experimental protocols and MS platforms. We will also gain a better understanding of the underlying mechanisms of non-specific binding - generating knowledge useful both for retrospective analysis of previously published data and for the design of future experiments. By integrating the experimental AP-MS data with external information such as structure-based protein interaction predictions, we will further improve the sensitivity of detection of low abundance and transient interactions. It has also become apparent that charting a complete interaction map for an organism like human is a community-wide effort, with multiple groups contributing separate portions of the entire interactome. We will develop a novel computational framework for consistent integration of AP-MS datasets from different studies, leading to more complete and accurate quantitative interaction networks. Lastly, one important problem that has yet to be fully addressed is the quantitative analysis protein complexes and interaction networks dynamics. The emergence of highly sensitive multiplex MS techniques presents such an opportunity, and we will develop advanced computational algorithms and tools for differential and dynamic interactome analysis using multiplex MS data. We will continue providing our widely used computational tools and data resources to the biological community, along with benchmark datasets for further development of computational methods by other scientists.