Calculating Dynamic Ensembles of Intrinsically Disordered Proteins

传统的结构-功能范式为折叠良好的蛋白质提供了重要的见解，其中 通过X射线晶体学光束线和NMR可以容易且快速地揭示结构。然而 大约三分之一的人类蛋白质组由内在无序的蛋白质和区域组成 不采用占主导地位的良好折叠结构，因此仍然被传统结构“看不见”， 生物学方法无序原子结构描述的实验和计算方法 蛋白质，虽然往往有价值，仍然缺乏预测能力，特别是对动态复合物的国内流离失所者，以及 缺乏对IDP结构集合与功能之间关系的深入了解。这是因为， 需要前所未有的多个和互补的基于解决方案的实验的集成水平， 国家的最先进的分子模拟，以提供现实的和相关的模型，为IDP合奏，选择 最好的合奏通过贝叶斯概率方法给定的欠定性质的问题， 综合分析，将观察到的动态结构与生物学相关的功能联系起来， 正在解决的问题。我们建议开发IDP计算器，它将（1）量化 大量实验数据类型的有用性和信息内容，例如化学位移、标量 耦合，RDC，NOE，PRE和FRET/FCS;（2）使用各种先进的原子和粗粒度 模型和抽样方法，用于产生IDP及其复合物的候选系综;（3）应用新的 用于IDP集成选择的贝叶斯模型，其评估并优化候选集成， 最好的实验数据类型;（4）创建一个软件套件，将这些方法与沿着 对结构、序列、绑定和其他功能数据进行相关分析的工具 国内流离失所者的问题。待开发的计算方法将推进结构的表征 具有内在紊乱的蛋白质的集合，不仅针对游离单体，而且强调IDP 配合物我们将在广泛的系统上开发和验证我们的方法： Sic 1：Cdc 4调节酵母细胞周期，蛋白磷酸酶1与其 控制不同细胞过程的无序调节因子，以及单体和相分离的FUS和 TDP-43 IDP对于理解导致生物相的动态分子间接触很重要 分离和ALS相关的聚合。

The traditional structure-function paradigm has provided significant insights for well-folded proteins in which structures can be easily and rapidly revealed by X-ray crystallography beamlines and NMR. However approximately one third of the human proteome are comprised of intrinsically disordered proteins and regions that do not adopt a dominant well-folded structure, and therefore remain “unseen” by traditional structural biology methods. Current experimental and computational approaches to structural descriptions of disordered proteins, while often valuable, still lack predictive power, particularly for dynamic complexes of IDPs, as well as lack of insight into the relationships between IDP structural ensembles and function. This is because IDPs require an unprecedented level of integration of multiple and complementary solution-based experiments, state-of-the art molecular simulations to provide realistic and relevant models for IDP ensembles, selection of the best ensembles via Bayesian probabilistic approaches given the underdetermined nature of the problem, and comprehensive analysis to connect observed dynamic structure with function relevant to the biological questions being addressed. We propose the development of IDP Calculator, which will (1) quantify the usefulness and information content of a large set of experimental data types such as chemical shifts, scalar couplings, RDCs, NOEs, PREs, and FRET/FCS; (2) use a variety of advanced atomistic and coarse-grained models and sampling methods for generating candidate ensembles of IDP and their complexes; (3) apply new Bayesian models for IDP ensemble selection that both evaluates and optimizes the candidate ensembles with the best experimental data types; and (4) create a software suite that will integrate these methods along with tools to perform correlative analysis of structural, sequence, binding and other functional data on a wide range of IDP problems. The computational approaches to be developed will advance the characterization of structural ensembles for proteins with intrinsic disorder, not only for the free monomer, but with emphasis on IDP complexes. We will develop and validate our approaches on a wide range of systems: the dynamic complex of Sic1:Cdc4 that regulates the yeast cell cycle, complexes formed between protein phosphatase 1 and its disordered regulators that control diverse cellular processes, and monomeric and phase-separated FUS and TDP-43 IDPs important for understanding the dynamic intermolecular contacts leading to biological phase separation and ALS-associated aggregation.