Development of a Next-Generation Nucleic Acid Force Field

项目摘要/摘要 核酸(Nas)是携带生命信息的主要分子。使用的能力 对DNA和RNA的结构、动态和相互作用进行建模的计算是一个关键的辅助工具 这些生物分子的实验研究。例如，含有伪尿嘧啶的mRNA 针对新冠肺炎的疫苗是抗击这种大流行的关键工具。DNA、mRNA和 MiRNAs是许多抗菌和抗病毒药物的靶标。小分子设计 与核酸结合的药物治疗癌症和神经退行性疾病 是当前药学研究中最热门的话题之一。在这个项目下，我们将 研究核酸的几个关键方面，发展可极化的多极 用于模拟Nas及其相互作用的阿米巴+模型。这个新的力场将进一步 通过将基于机器学习的局部价态特征的潜力与 经典的长程非键相互作用。由此产生的变形虫+NN力场有望 NA体系结合计算中的化学准确度。阿米巴+的参数 而阿米巴+NN势将使用新的自动化Poltype2组件产生。 力场在现有的支持GPU的Tinker9分子动力学中的实现 软件将实现最先进的模拟和束缚自由能估计。 分子模拟在治疗学设计中的适用性受到效率和 计算的准确性。这项建议的目的是使常规、准确 小分子配体与NaS结合的热力学计算。朝向那个方向 最后，将调查几个当前的生物相关性系统，包括结合 金属离子与G-四链结构、带有RNA适配子的荧光配体、新型抗生素 药物与FMN核糖开关，以及P5abc结构域的构象动力学 四膜虫I群核酶。Nas的结构和功能高度依赖于 它们的离子环境。RNA局部结构动力学与全球/第三纪的相互作用 折叠是一个正在实验中解决的有趣问题。能够模拟这些 新型小分子药物设计中的复杂能量效应及其合成修饰 寡核苷酸将是未来医学发展的一个重要增长领域。发展中的 精确且可移植的下一代阿米巴+和阿米巴+NN力场将 为理解和预测自然和设计的行为开辟了新的途径 核酸分子。

PROJECT SUMMARY/ABSTRACT Nucleic acids (NAs) are the major information carrying molecules of life. The ability to use computation to model the structure, dynamic and interactions of DNA and RNA is a key adjunct to experimental study of these biomolecules. For example, pseudouridine-containing mRNA vaccines against Covid-19 are a critical tool in battling the pandemic. DNAs, mRNAs, and miRNAs are targets for a number of antibacterial and antiviral drugs. Design of small molecule drugs binding to nucleic acids in the treatment of cancers and neurodegenerative diseases is one of the hottest topics of current pharmaceutical research. Under this project, we will investigate several key aspects of nucleic acids, and develop the polarizable multipole AMOEBA+ model for simulation of NAs and their interactions. This new force field will be further enhanced via coupling of a machine learning-based potential for local valence features with classical long-range nonbonded interactions. The resulting AMOEBA+NN force field promises chemical accuracy in the calculation of binding for NA systems. Parameters for the AMOEBA+ and AMOEBA+NN potentials will be generated using the new, automated Poltype2 package. Implementation of the force fields into the existing GPU-capable Tinker9 molecular dynamics software will enable state-of-the-art simulation and binding free energy estimation. The applicability of molecular simulation to design of therapeutics is limited by efficiency and accuracy of the calculations. The objective of this proposal is to enable routine, accurate computation of the thermodynamics of binding of small-molecule ligands to NAs. Toward that end, several current systems of biological relevance will be investigated, including binding of metal ions to G-quadruplex structures, fluorogenic ligands with RNA aptamers, novel antibiotic drugs with the FMN riboswitch, and conformational dynamics of the P5abc domain of the Tetrahymena group I ribozyme. The structures and functions of NAs are highly dependent upon their ionic environment. The interplay between RNA local structural dynamics and global/tertiary folding is an intriguing question being addressed experimentally. The ability to simulate these complex energetic effects in the design of novel small molecule drugs and synthetically modified oligonucleotides will be an important growth area for future medical advances. Development of the accurate and transferable next-generation AMOEBA+ and AMOEBA+NN force fields will open new paths toward understand and prediction of the behavior of natural and designed nucleic acid molecules.