RUI: New Tools for Characterizing Protein Dynamics

位于默塞德的加州大学获得了一笔赠款，用于开发表征蛋白质动力学的新工具和协议。虽然蛋白质生物物理学的重点传统上是天然折叠的蛋白质的结构和相互作用，现在有迅速增长的兴趣变性蛋白质结构的性质，有关的蛋白质折叠的机制和所谓的内在无序蛋白质的功能。分子模拟在帮助理解无序蛋白质方面有很大的希望;然而，需要新的工具来表征这些不断变化的系统的定义属性-它们的动力学。这项研究的长期目标是开发新的工具和协议，用于确定蛋白质动力学的两个最基本的描述符，运动的有效维度和蛋白质采样的结构空间的体积。这些工具将有助于回答许多关于蛋白质变性或无序状态的基本的、未解决的问题，这些问题对于理解蛋白质功能和折叠至关重要。具体而言，这些工具将确定变性或内在无序的蛋白质在多大程度上表现为随机聚合物或具有受限于低维动力学的运动。此外，这些工具将阐明是否有内在无序的蛋白质和天然折叠的蛋白质在折叠的早期阶段的动力学之间的差异。最后，这些工具将有助于确定不同的分子模拟在采样蛋白质的变性或无序状态的效率。该项目将为参与的学生提供理想的多学科培训体验，因为它汇集了三个科学学科：机器学习，分子模拟和聚合物生物物理学。在这个项目中开发的工具将用于研究生和本科生在分子模拟和计算机科学课程在加州大学默塞德和学生在夏季研究项目。分子动力学社区有一个良好的软件工具的传统，这些工具由各个研究小组开发和维护，并免费分发给非常大的用户群。这些已建立的渠道将用于以源代码格式沿着文档和测试数据集免费分发本项目中开发的所有软件工具。更多信息可以在项目网站上找到：http://vision.ucmerced.edu/projects/protein。

The University of California at Merced has been awarded a grant to develop new tools and protocols for characterizing protein dynamics. While the focus of protein biophysics has traditionally been on the structure and interactions of the natively folded protein, there is now rapidly growing interest in the properties of denatured protein structures as related to the mechanisms of protein folding and in the function of so-called intrinsically disordered proteins. Molecular simulations have great promise in helping to understand disordered proteins; however, new tools are needed to characterize the defining property of these constantly changing systems--their dynamics. The long-term goal of this research is to develop new tools and protocols for determining the two most fundamental descriptors of protein dynamics, the effective dimensionality of the motion and the volume of structural space sampled by the protein. The tools will help answer the many fundamental, unresolved questions about the denatured or disordered states of proteins that are central to understanding protein function and folding. Specifically, these tools will determine to what extent denatured or intrinsically disordered proteins behave as random polymers or have motions constrained to lower dimensional dynamics. Additionally these tools will elucidate whether there are differences between the dynamics of intrinsically disordered proteins and natively folded proteins in the early stages of folding. Finally, these tools will help in determining the efficiency of different molecular simulations in sampling the denatured or disordered states of proteins. This project will be an ideal multidisciplinary training experience for the students involved since it brings together three scientific disciplines: machine learning, molecular simulation, and polymer biophysics. The tools developed in this project will be used by graduate and undergraduate students in both molecular simulation and computer sciences courses at UC Merced and by students in summer research projects. The molecular dynamics community has a well established tradition of software tools that are developed and maintained by individual research groups and freely distributed to very large user bases. These established channels will be used to freely distribute all of the software tools developed in this project in source code format along with documentation and test data sets. More information can be found at the project website: http://vision.ucmerced.edu/projects/protein.