Computational tools for mass spectrometry-based interactome data

描述(由申请人提供)：在这里，我们建议扩展我们正在进行的项目，开发使用亲和纯化质谱学技术(AP-MS)分析和模拟蛋白质-蛋白质相互作用(PPI)数据的计算方法和工具。最近的技术进步使AP/MS成为体内研究PPI的一种广泛使用的技术。与此同时，针对这些数据的计算方法和工具的开发也相对滞后。对于PPI数据的统计评估，迫切需要准确和稳健的方法。此外，AP/MS数据目前被用于重建纯粹定性的蛋白质网络。一些重要的问题，如蛋白质在多个复合体中的分配，以及在细胞环境中的扰动下复合体组成的变化等，仍有待于充分解决。作为我们正在进行的R01项目的一部分，我们正在努力通过利用MS数据中编码的定量信息，为基于MS的PPI网络和复合体的分析增加一个新的维度。目前，我们专注于常用的无标记定量策略，如光谱计数和多肽母体离子强度。利用这些数据，我们已经证明了我们的方法能够更准确地重建蛋白质复合体和相互作用网络。我们还开发并继续改进了一个统计框架，即相互作用组的显著性分析(SANT)，它利用光谱计数或多肽强度来为个体相互作用分配置信度度量。我们正在开发的用于PPI数据的统计分析和数据挖掘的SAINT和其他工具已经被世界各地越来越多的实验室使用。然而，最近出现了几种新的策略，它们都是基于基于MS的靶向蛋白质定量的概念。特别是，新的被称为SWATH-MS的MS方法为开发更快、更准确和高灵敏度的PPI监测技术提供了一个令人兴奋的机会。将SWATH-MS方法与亲和纯化AP/SWATH-MS相结合，应该使我们能够以非常高的精度确定给定AP/MS数据集中每个已识别的相互作用蛋白质的相对丰度。总而言之，这将为以动态方式分析PPI和网络创造一个技术平台。它还将提高我们监测涉及特定蛋白质形式的PPI的能力，包括翻译后修饰(例如磷酸化)的蛋白质。按照我们正在进行的研究合作的总体主题，AP/SWATH-MS方法将在几项生物学研究的背景下进行优化和评估，这些研究通过它们对细胞信号的基本理解的重要性而联系在一起。我们合作研究的最终目标是使用AP/MS及其后续的生物解释来生成高质量的PPI网络和复合体。将这项工作扩展到包括新兴的SWATH-MS技术，有可能显著提高PPI网络重建的准确性和吞吐量，并应能够以动态方式有效监测相互作用网络的变化。

DESCRIPTION (provided by applicant): Here we are proposing to extend our ongoing project on the development of computational methods and tools for analyzing and modeling protein-protein interaction (PPI) data using affinity-purification mass spectrometry technology (AP-MS). The recent technological advances have made AP/MS a widely used technique for studying PPIs in vivo. At the same time, the development of computational methods and tools for these data has lagged behind. There is a great need for accurate and robust methods for statistical assessment of PPI data. Furthermore, AP/MS data is presently used to reconstruct protein networks that are purely qualitative in nature. Such important questions as partition of proteins into multiple complexes and changes in the complex composition upon perturbation in the cellular environment remain to be fully addressed. As a part of our ongoing R01 project, we are working to add a new dimension to MS-based analysis of PPI networks and complexes by taking advantage of the quantitative information encoded in MS data. At the moment, we are focusing on commonly used label-free quantification strategies such as spectral counting and peptide parent ion intensities. Using these data, we have already demonstrated that our methods enable more accurate reconstruction of protein complexes and interaction networks. We have also developed and continue improving a statistical framework, Significance Analysis of Interactome (SAINT), which utilizes spectral counts or peptide intensities for assigning a confidence measure to individual interactions. SAINT and other tools we are developing for statistical analysis and data mining of PPI data are already being used by an increasing number of laboratories worldwide. Very recently, however, several new strategies have emerged that are all based on the concept of targeted MS-based protein quantification. In particular, the new MS approach called SWATH-MS offers an exciting opportunity for developing a much faster, more accurate, and highly sensitive technology for monitoring PPIs. Coupling SWATH-MS approach with affinity purification, AP/SWATH-MS, should enable us to determine - with very high accuracy - the relative abundance of every identified interacting protein in a given AP/MS dataset. This, in term, will create a technological platform for the analysis of PPIs and networks in a dynamic fashion. It will also improve our ability to monitor PPIs involving specific forms of proteins, including post-translationally modified (e.g. phosphorylated) proteins. Following the overall theme of our ongoing research collaboration, the AP/SWATH-MS approach will be optimized and evaluated in the context of several biological studies linked through their significance for fundamental understanding of cell signaling. The ultimate goal of our collaborative research is to enable generation of high quality PPI networks and complexes using AP/MS and their subsequent biological interpretation. Extending this work to include the emerging SWATH-MS technology has a potential to significantly improve the accuracy and throughput of PPI network reconstruction, and should enable effective monitoring of changes in the interaction networks in a dynamic fashion.