Theory and Modeling of Noncovalent Binding

许多药物是小的有机分子，它们结合了一种酶或受体，这种酶或受体参与了一种 疾病过程。化学宿主代表了另一类具有生物医学意义的化合物；这些化合物是 比蛋白质小得多，但仍然有裂隙的分子，使它们能够结合 靶向配体。化学宿主用于提高药物的生物利用度和稳定性，以及 有潜在的新用途作为药品本身；例如，作为毒素清除剂或 即使是作为人造酶。目前，还没有可靠的计算方法来设计靶标 配体或化学主体，因此此类化合物的发现在很大程度上依赖于昂贵和时间- 耗费大量的实验性试验和错误。 这项提议旨在通过开发出一种新的方法来加速发现用于医学应用的靶向化合物 改进的理论和计算方法，包括自动化方法来设计 化学宿主，以及一种计算配体-蛋白质结合亲和力的准确新方法。这个 这里采用的方法集中在主要状态模型上，这些模型在 最近在PI实验室的主-客系统中的应用。这些方法高度可并行化， 该项目将受益于超级计算专业知识和合作中可用的资源 梅奥诊所医学院的实验室。

Many medications are small organic molecules that bind an enzyme or receptor that is involved in a disease process. Chemical hosts represent another class of compounds of biomedical interest; these are molecules which are much smaller than proteins but which still possess a cleft enabling them to bine1 a targeted ligand. Chemical hosts are used to improve the bioavailability and stability of medications, and have potential new uses as Pharmaceuticals in their own right; for examples as scavengers of toxins or even as artificial enzymes. Currently, there is no reliable computational method for designing targeted ligands or chemical hosts, so the discovery of such compounds relies heavily upon costly and time- consuming experimental trial and error. This proposal aims to speed the discovery of targeted compounds for medicalapplications by developing improved theoretical and computational methods, including ah automated approach to the design of chemical hosts, and an accurate new method of computing ligand-protein binding affinities. The approaches taken here focus on predominant states models, which have provided promising results in recent applications to host-guest systems in the Pi's laboratory. These methods are highly parallelizable, and this project will benefit from supercomputing expertise and resources available in the collaborating laboratory at the Mayo Clinic School of Medicine.