Equipment Supplement to R35GM146987: Purchase of LC-MS system for high throughput isolation of bioactive natural products

项目概要/摘要 来自细菌、真菌和植物的天然产物长期以来一直是具有令人着迷的分子的丰富来源 化学结构和治疗相关的生物活性。但由于其结构复杂， 很难筛选许多天然产品的类似物以真正理解控制关系的规则 它们的结构和活动之间。在家长奖中，我们通过开发来应对这一挑战 机器学习方法可以对自然的结构-活动关系（SAR）进行功能建模 产品并有助于设计可以合成天然产物类似物的生物合成途径。 家长奖的第一个项目应用机器学习来研究天然产物的SAR。我们的一种方法是 采用的是一种遗传方法，我们正在验证我们之前开发的机器学习模型 根据天然产物生物合成中是否存在生物合成基因来预测生物活性 基因簇（BGC）。这种方法的一个主要挑战是细菌培养物的提取会产生复杂的 代谢物的混合物，很难将单个代谢物与其 BGC 或观察到的活性联系起来。目前， 我们正在使用标准的生物活性引导分馏将天然产物与观察到的活性联系起来。这是一个低 吞吐量和费力的过程。有些技术可以使这个过程更快，通常 涉及并行执行色谱法并收集馏分和质谱，以便比质量 特征可以与每个分数相关联。然后，在生物活性测定中观察到的活性可以与 特定的质量特征来识别哪些分子负责活性。这个过程要求比较高 通量比我们当前的方法高，但它需要具有馏分收集功能的 LC-MS，而我们没有 使用最新的仪器来进行该测定。该设备补充将使我们能够 购买所需的仪器并最大限度地提高该项目成功的机会。 家长奖的第二个项目专注于开发机器学习和其他计算工具 用于设计 BGC 来生物合成新型天然产物类分子。在本项目中，我们将使用 LC-MS 来 确定我们的工程生物合成酶是否产生预期的产物。虽然我们确实可以访问 适用于此目的的核心设施工具，使用它们需要按小时付费。我们 也可以使用我们随本补充品购买的仪器来进行这些实验，这将 节省核心设施成本，从而使更多的家长奖励用于人员和用品。

Project Summary/Abstract Natural products from bacteria, fungi, and plants have long been a rich source of molecules with fascinating chemical structures and therapeutically-relevant bioactivities. However, due to their complex structures, it is difficult to screen many analogs of natural products to truly understand the rules governing the relationship between their structure and activity. In the parent award, we are addressing this challenge by developing machine learning methods that can functionally model the structure-activity relationships (SAR) of natural products and aid in the design of biosynthetic pathways that can synthesize natural product analogs. The first project of the parent award applies machine learning to study natural product SARs. One approach we take is a genetic approach, where we are validating a machine learning model we previously developed that predicts bioactivity based on the presence or absence of biosynthetic genes in a natural product’s biosynthetic gene cluster (BGC). One major challenge of this approach is that extractions of bacterial cultures yield a complex mixture of metabolites and it is difficult to link individual metabolites to their BGCs or observed activity. Currently, we are using standard bioactivity-guided fractionation to link natural products to observed activity. This is a low- throughput and laborious process. There are technologies that can make this process faster, which generally involve performing chromatography and collecting fractions and mass spectra in parallel, so that specific mass features can be associated with each fraction. Then activity observed in bioactivity assays can be correlated with specific mass features to identify which molecules are responsible for activity. This process is much higher throughput than our current approach but it requires an LC-MS with fraction collection, and we do not have access to an up-to-date instrument for performing this assay. This equipment supplement will allow us to purchase the required instrument and maximize the chances that this project will be successful. The second project of the parent award focuses on developing machine learning and other computational tools for designing BGCs to biosynthesize novel natural product-like molecules. In this project, we will use LC-MS to determine if our engineered biosynthetic enzymes produced the expected product. While we do have access to core facility instruments that are appropriate for this purpose, using them is associated with an hourly fee. We could also use the instrument we will purchase with this supplement to perform these experiments, which would save on core facility costs allowing for more of the parent award to go to personnel and supplies.