SI2-SSE: Enhanced Software Tools for Biomolecular Free Energy Calculations

基于原子模拟的自由能计算在基础化学和生物研究中的核心作用现已牢固确立。自由能是指导处于或接近平衡状态的系统行为的核心量，它决定了分子构象、分子结合、化学反应等特征。不幸的是，准确可靠的自由能很难计算，特别是对于许多以崎岖的自由能景观为特征的生物分子系统。因此，需要特殊的技术来计算这种自由能景观。之前已经开发了所谓的自适应偏倚分子动力学方法，并通过其他方法（由我们小组开发或改编）增强了相空间采样，该项目将进一步开发该软件的功能，从而扩大可以解决的模拟问题的种类。该软件将作为开源软件和AMBER软件包的一部分向公众发布。在科学应用方面，本项目还将研究具有内在无序区域的蛋白质的构象和性质，以及DNA与一类特殊转录因子的结合。描述生物分子和化学平衡系统的最重要的一个量是自由能。然而，计算自由能是出了名的困难和昂贵的计算。这个问题对于许多生物分子系统来说是特别紧迫的，这些系统的特点是复杂的自由能景观，很难用常规的分子动力学模拟来探索。pi先前开发了具有多步行者的自适应偏置分子动力学方法（ABMD）和副本交换分子动力学（REMD）扩展。ABMD是一种伞式采样方法，具有时变偏置电位，可用于计算自由能景观和构象采样。该软件套件，连同导向分子动力学（SMD）扩展，已经作为AMBER软件包的一部分向公众发布。该项目将通过开发处理基于四元数的集体变量的能力，即所谓的里程碑技术，以及自我导向、相互作用的多步行者，将这套软件工具提升到一个新的水平。预期的应用涉及具有内在无序区（IDRs）的蛋白质和转录因子中基本螺旋-环-螺旋（bHLH）结构域的DNA结合机制。

The central role of atomistic simulation-based free energy calculations for basic chemical and biological research is now firmly established. The free energy is the central quantity that guides the behavior of a system at, or near, equilibrium, determining such characteristics as molecular conformations, molecular binding, chemical reactions, etc. Unfortunately, accurate and reliable free energies are very difficult to calculate, particularly for many biomolecular systems characterized by rugged free energy landscapes. Hence, special techniques are required for calculating such free energy landscapes. Having previously developed the so-called Adaptively Biased Molecular Dynamics method, enhanced with other methods (either developed or adapted by our group) for phase space sampling, this project will further develop the capabilities of this software thereby enlarging the kinds of simulation problems that can be tackled. The software will be released to the public as open source software and as parts of the AMBER software package. In terms of scientific applications, this project also will investigate the conformation and properties of proteins with Intrinsically Disordered Regions, and the binding of DNA with a special class of transcription factors.The single most important quantity for describing biomolecular and chemical systems in equilibrium is the free energy. However, calculating free energies is notoriously difficult and computationally expensive. This problem is particularly pressing for many biomolecular systems, which are characterized by complicated free energy landscapes that are hard to explore with regular molecular dynamics simulations. The PIs previously developed the Adaptively Biased Molecular Dynamics method (ABMD) with Multiple Walkers, and Replica Exchange Molecular Dynamics (REMD) extensions. ABMD is an umbrella sampling method with a time-dependent biasing potential for calculating free energy landscapes and conformational sampling. The software suite, along with Steered Molecular Dynamics (SMD) extensions, has been released to the public as part of the AMBER software package. This project will take this set of software tools to the next level by developing the capability to handle quaternion-based collective variables, the so-called Milestoning technique, and self-directed, interacting multiple walkers. Envisioned applications relate to proteins with Intrinsically Disordered Regions (IDRs) and mechanisms of DNA binding by basic Helix-Loop-Helix (bHLH) domains in transcription factors.