Tools for rapid and accurate structure elucidation of natural products

应用摘要 5 R01 GM107550-08 “快速准确的工具 天然产物的结构解析” 细菌是结构独特且具有生物活性的天然物质的极其丰富的来源 源自多种令人着迷的生化途径的产品。然而，完整的 天然产物的结构解析通常是最耗时且成本最高的工作 天然产物药物发现计划。更复杂的是，基因组的进步 测序加速了独特的模块化生物合成基因簇的鉴定 原核生物并揭示了大量尚未分离和生物学的新化合物 化学表征。因此，该领域迫切且持续需要 将生物合成基因簇连接到 MS2 中各自的 MS 碎片特征 分子网络。建立这种连接的能力将加速新化合物的开发 发现以及在基因簇和生物合成途径之间建立关联，并帮助 快速准确的结构解析。结合这种信息学方法， 拟议的延续项目探索解决复杂分子问题的创新方法 通过增强的 MS 和 NMR 实验以及新算法的开发来确定结构 通过它来加速他们的分析。因此，这笔赠款的总体目标是开发高效的 促进自动结构分类、结构特征发现和 最终有效地阐明天然产物（或任何小分子）的结构并构建 与社区输入的数据进行交互的基础设施。我们将通过以下方式实现这一目标 以下四个具体目标： 目标 1. MS2 分子网络与基因的整合 集群网络快速高效地定位具有独特性的天然产品 分子结构；目标 2. 开发一套高灵敏度脉冲序列 天然产物结构阐明；目标 3. 开发基于 NMR 的分子网络 使用深度卷积神经网络（DCNN）的策略来促进 有机化合物的分类和结构解析；目标 4. 整合 NMR 分子网络和基于 MS2 的分子网络作为有效的结构 表征和阐明策略。通过实现这些目标，我们将开发一个 用于快速寻找新化合物并确定其结构的创新工作流程 并准确地。基因簇和分子之间的联系将揭示 立体化学和潜在的卤化和甲基化。然后可以使用该信息 结合更有效的 NMR 和 MS 方法来准确确定结构。 这些工具将被广泛共享，例如通过全球天然产品社交 (GNPS) 分子网络，提升天然产物和有机化学综合能力 社区来解决复杂的分子结构。

Summary from the Application 5 R01 GM107550-08 “Tools for rapid and accurate structure elucidation of natural products” Bacteria are extraordinarily prolific sources of structurally unique and biologically active natural products that derive from a diversity of fascinating biochemical pathways. However, the complete structure elucidation of natural products is often the most time consuming and costly endeavor in natural product drug discovery programs. Compounding this, advancements in genome sequencing have accelerated the identification of unique modular biosynthetic gene clusters in prokaryotes and revealed a wealth of new compounds yet to be isolated and biologically and chemically characterized. Resultantly, there is an urgent and continuing need in this field to connect biosynthetic gene clusters to their respective MS fragmentation signatures in the MS2 molecular networks. The capacity to make such connections will accelerate new compound discovery as well as create associations between gene cluster and biosynthetic pathway, and aid in fast and accurate structure elucidations. Combined with this informatics approach, this proposed continuation project explores innovative methods by which to solve complex molecular structures by enhanced MS and NMR experiments, as well as the development of new algorithms by which to accelerate their analysis. Thus, the overarching goal of this grant is to develop efficient methods that facilitate automated structural classification, structural feature discovery and ultimately efficient structure elucidation of natural products (or any small molecule) and to build an infrastructure that interacts with data input from the community. We will achieve this with the following four specific aims: Aim 1. Integration of MS2 molecular networking with gene cluster networking to rapidly and efficiently locate natural products that have unique molecular architectures; Aim 2. To develop a suite of high sensitivity pulse sequences for natural product structure elucidation; Aim 3. To develop NMR based molecular networking strategies using Deep Convolutional Neural Networks (DCNNs) to facilitate the categorization and structure elucidation of organic compounds; Aim 4. To integrate NMR molecular networking and MS2 -based molecular networking as an efficient structure characterization and elucidation strategy. By achieving these aims we will develop an innovative workflow for finding new compounds and for determining their structures, both quickly and accurately. The connection between gene cluster and molecule will shed light on stereochemistry and potential halogenations and methylations. This information can then be used in combination with more efficient NMR and MS methods to accurately determine structures. These tools will be widely shared, such as through the Global Natural Products Social (GNPS) Molecular Network, to enhance the overall capacity of the natural products and organic chemistry communities to solve complex molecular structures.