Theory and Modeling of Noncovalent Binding

描述(申请人提供)：拟议的研究，纯粹是理论和计算，有三个主要部分，都集中在分子识别现象的各个方面，这在生物学和医学中具有基础和实际意义。第一部分涉及被称为宿主的微型受体分子，它们今天被用来稳定和输送许多药物，它们本身也显示出作为药物的前景。我们计划向其他科学家提供新的、经过测试的软件，帮助他们设计这些微型受体，从而加快新药的开发。我们还计划对这些分子进行模拟，以便更好地了解它们是如何工作的，并了解更大的受体，如许多蛋白质，是如何与药物结合的。第二部分是研究分子结合时发生的熵变化。在最近的工作中，我们发现与受体及其结合的分子(配体)的运动相关的熵的变化可以对它们相互结合的紧密程度产生令人惊讶的强烈影响。然而，在设计紧密结合的受体和配体时，我们还不能很好地理解这些熵变化，使它们对我们有利。我们计划进一步发展我们通过计算机模拟计算这些熵变化的方法，然后使用该方法来更好地理解它们。例如，我们希望能够预测何时修改一个配体，使其更刚性，从而降低熵，将增加其亲和力。此外，我们计划将新的熵计算纳入计算结合亲和力的软件中，我们希望这将有助于研究人员设计新药。第三部分提出了将应力概念应用于分子生物物理学的新思路。材料科学家已经通过原子计算机模拟得出了计算材料中应力的公式，我们认为这些公式可以告诉我们一些关于宿主、蛋白质和其他分子如何工作的有用信息。首先，我们假设，如果受体与配体结合产生局部应力，那么修改配体以降低这种应力可能会增加结合亲和力。因此，计算应力可能有助于设计紧密结合的配体。我们还假设，当配体结合变构蛋白时，变构蛋白的构象在结合时发生变化，形状变化的机制涉及从结合部位传播应力波。如果我们能够了解蛋白质是如何改变构象的，这将有助于我们重新设计它们，用于医疗和工业用途。 公共卫生相关性：该项目将化学理论和计算机建模应用于分子识别现象。我们的总体目标是更好地理解是什么让特定的分子相互结合，并将这种理解融入到软件中，这将在蛋白质工程和新药设计中有用。

DESCRIPTION (provided by applicant): The proposed research, which is purely theoretical and computational, has three main parts, all focused on aspects of the phenomenon of molecular recognition, which is of basic and practical importance in biology and medicine. The first part concerns miniature receptor molecules, called hosts, which are used today to stabilize and deliver many drugs and which also show promise as medications in their own right. We plan to provide other scientists with new, tested software to help them design these miniature receptors and thereby speed the development of new medications. We also plan to carry out simulations of these molecules in order to develop a better understanding of how they work and also to gain insight into how larger receptors, such as many proteins, bind drugs. The second part is to study the changes in entropy that occur when molecules bind. In recent work, we have found that changes in entropy associated with the motions of receptors and the molecules they bind (ligands) can have a surprisingly strong influence on how tightly they bind each other. However, we do not yet understand these entropy changes well enough to make them work in our favor when designing tight-binding receptors and ligands. We plan to further develop our method of computing these entropy changes from computer simulations, and then use the method to develop a better understanding of them. For example, we would like to be able to predict when modifying a ligand to make it more rigid, and therefore lower in entropy, will increase its affinity. In addition, we plan to incorporate the new entropy calculations into software for computing binding affinities which we hope will help researchers design new drugs. The third part is to develop a new idea of applying the concept of stress to molecular biophysics. Materials scientists have come up with equations for computing the stress in a material from an atomistic computer simulation, and we think these equations can tell us something useful about how hosts, proteins and other molecules work. For one thing, we hypothesize that if receptor- ligand binding produces localized stress, then modifying the ligand to reduce this stress might increase the binding affinity. Thus, computing stress might help with the design of tight-binding ligands. We also hypothesize that, when a ligand binds an allosteric protein, a protein whose conformation shifts on binding, the mechanism of the shape change involves propagation of a wave of stress from the binding site. If we can understand how proteins change conformation, this would help us to re-engineer them for medical and industrial uses. PUBLIC HEALTH RELEVANCE: This project applies chemical theory and computer modeling to the phenomenon of molecular recognition. Our overall goal is to develop a better understanding of what makes specific molecules bind each other, and to incorporate this understanding into software that will be useful in protein engineering and the design of new medications.