ABI Innovation: A Metabolomics Toolkit for NMR and Mass Spectrometry

在生物基础设施部生物信息学进展的支持下，内布拉斯加州大学林肯分校的Robert Powers教授及其研究小组将进一步开发他们的MVAPACK软件工具包，用于分析代谢组学研究中产生的数据。代谢组学是研究小分子的化学特征，如糖，是生物过程的产物。代谢物指纹图谱揭示了生物样品中近期发生的大量活动。了解生物体如何对周围环境做出反应，对于理解植物生物学和农业、环境和水质研究以及营养学或疾病方面的人类健康研究等领域的事件具有重要意义。化学指纹是通过两种不同的技术生成的，分别是核磁共振和质谱。结果可能非常复杂且难以合并。这意味着数据分析的挑战也很高，包括开发新方法的要求，如多元统计方法。本提案旨在开发软件工具，从数据中提取最可能的信息，包括数据格式转换、预处理、验证、显著性检验、可视化和确定置信度水平的代谢物鉴定。MVAPACK工具包将以免费和开源软件模块的形式实现重要的、目前不可用的数据处理功能，对数据质量、共享、信息内容产生影响，从而从包括代谢物分析在内的研究中获得更深入的生物学见解。在授奖期间，该研究项目将进一步增强MVAPACK代谢组学工具包，包括额外的和迫切需要的功能。这些改进将包括增加加载和处理液相色谱(LC)-质谱、气相色谱(GC)-质谱和多维数据集的功能；扩展能力，以整合化学计量学的多个数据集；自动化代谢物鉴定，这是代谢组学中最具挑战性和耗时的方面之一。这项工作将采用图论、协方差核磁共振、反向缩放伪光谱加载以及核磁共振和质谱相结合的方法来促进复杂生物样品中代谢物的准确和快速分配。总的来说，这些努力将为代谢组学科学界带来一个改进的MVAPACK软件包（http://bionmr.unl.edu/mvapack.php），这将增强信息内容，实用性，以及处理和组合各种类型的NMR和MS数据集的便利性。

With support from the Advances in Biological Informatics in the Division of Biological Infrastructure, Professor Robert Powers and his research group at the University of Nebraska - Lincoln will further develop their MVAPACK software toolkit that analyzes data generated in metabolomics research. Metabolomics is the study of the chemical signatures of the small molecules, like sugars, that are the products of biological processes. Metabolite fingerprints reveal a great deal about the recent activities occurring in biological samples. Knowing how organisms respond to their surroundings gives important leads to understanding events in fields like plant biology and agriculture, environmental and water quality studies, and human health research into nutrition or diseases. The chemical fingerprints are generated using two different technologies, called nuclear magnetic resonance and mass spectrometry. The results can be very complex and difficult to merge. This means the challenges for data analysis are also high, and include requirements to develop new approaches, like multivariate statistical methods. This proposal aims to produce software tools that will extract the most possible information from the data, including data format conversion, preprocessing, validation, significance testing, visualization, and metabolite identification at defined levels of confidence. The MVAPACK toolkit will enable important, currently unavailable data processing functions in the form of freely available and open source software modules, with an impact on data quality, sharing, information content resulting in deeper biological insights from studies that include metabolite profiling. During the award period, this research project will further enhance the MVAPACK metabolomics toolkit by including additional and urgently needed capabilities. Such improvements will include adding features to load and process liquid chromatography (LC)-MS, gas chromatography(GC)-MS and multidimensional datasets; to expand capabilities to integrate multiple datasets for chemometrics; and to automate metabolite identification, which is one of the more challenging and time-consuming aspects of metabolomics. This work will employ, among other approaches, graph theory, covariance NMR, back-scaled pseudospectral loadings, and a combined NMR and MS approach to facilitate the accurate and rapid assignment of metabolites from complex biological samples. Overall, these efforts will lead to an improved MVAPACK software package (http://bionmr.unl.edu/mvapack.php) for the metabolomics scientific community that will enhance the information content, utility, and ease of processing and combining various types of NMR and MS datasets.