Computational Core

项目摘要-计算核心 西海岸代谢组学化合物鉴定中心（WCMC）的计算核心是 致力于NIH共同基金代谢组学倡议的总体目标，特别是旨在大大 改进小分子鉴定。计算核心领导人Kind博士和量子化学专家Prof. Tantillo和王教授将与项目主任Fiehn教授以及计算专家和程序员一起工作 具体目标如下。1)生成用于GC-MS的大型计算机质谱库 和LC-MS/MS应用程序，以改善代谢组学中的化合物注释。这些计算机光谱库 将通过使用量子化学，包括玻恩-奥本海默分子动力学和跃迁， 状态反应建模。包括经典方法，如启发式和机器学习模型， 从计算预测的酶混杂数据库中衍生的新化合物。2)的 开发高精度的计算机碎片集成，用于未知质谱的快速排序。 这种方法在还没有参考光谱可用但化合物被覆盖的情况下是有利的。 在现有的化学数据库。这包括开发一个过滤管道，该管道使用从 WCMC实验核心，并纳入额外的先验数据和参考文献。3)发展 在GC-MS和LC-MS中为各种各样的大量化合物提供高度准确的保留预测方法 用于化合物鉴定报告的正交过滤和细化。软件工具和 数据库将与指导委员会和 行政核心。

Project Summary – Computational Core The Computational Core at the West Coast Metabolomics Center for Compound Identification (WCMC) is committed to the overall goals of the NIH Common Fund Metabolomics Initiative and specifically aims to greatly improve small molecule identifications. Computational Core leader Dr. Kind and quantum chemistry experts Prof. Tantillo and Prof. Wang will work with Program director Prof. Fiehn and computational experts and programmers on the following specific aims. 1) The generation of large in-silico mass spectral libraries for use in both GC-MS and LC-MS/MS applications to improve compound annotations in metabolomics. These in-silico spectral libraries will be created by using quantum chemistry including Born-Oppenheimer molecular dynamics and transition state reaction modelling. Inclusion of classical approaches such as heuristic and machine learning models using novel compounds that are derived from computationally predicted enzyme promiscuity databases. 2) The development of high accuracy in-silico fragmentation ensembles, for fast ranking of unknown mass spectra. Such an approach is advantageous in case no reference spectrum is available yet, but the compound is covered in existing chemical databases. This includes the development of a filtering pipeline with data obtained from the WCMC Experimental Core and inclusion of additional priori data and literature references. 3) The development of highly accurate retention prediction methods for a diverse and large set of compounds in GC-MS and LC-MS to be used for orthogonal filtering and refinement of compound identification reports. The software tools and databases will be independently validated and tested in close collaboration with the steering committee and the Administrative Core.