Department of Bioinformatics, Institute of Biochemistry and Biophysics, University of Tehran, Tehran, Iran.
Department of Cell and Molecular Biology, School of Biology, College of Science, University of Tehran, Tehran, Iran.
J Mol Graph Model. 2019 May;88:183-193. doi: 10.1016/j.jmgm.2019.01.009. Epub 2019 Jan 24.
Structural characterization of intrinsically disordered proteins (IDPs) is paramount and challenging in structural biology. In this regard, a de novo computational protocol is introduced to build heterogeneous structural libraries for amyloid-β (Aβ) as a critical IDP. This method combines the strength of the simulated annealing - in jumping over energy barriers and escaping from traps - with short conventional molecular dynamics simulations to quickly explore local regions of the conformational space. The protocol efficiency and reliability in building Aβ conformational library is compared with two widely used simulation methods, replica exchange molecular dynamics and multiple trajectory sampling. The probability distribution functions of various structural and energetic features are constructed for each library, and also the diversity and convergence rates in these protocols were compared. Our results show that the suggested protocol is a successful computational method in the generation of a diverse conformational library of the Aβ monomer in agreement with experimental data. This method focuses on visiting more conformations in less computational time without paying attention to the statistical weight of each state in the library. We believe that the suggested computational technique can be used for generating a reasonable starting pool for subsequent reweighting with experimental data to obtain a statistical ensemble.
结构生物学中，对无规卷曲蛋白质（IDPs）进行结构特征分析至关重要，但极具挑战性。在这一方面，我们提出了一种从头计算的方案，以用于构建淀粉样蛋白-β（Aβ）的异质结构文库，Aβ 是一种关键的 IDP。该方法结合了模拟退火的优势——跳过能量势垒并摆脱陷阱——以及短时间的常规分子动力学模拟，从而快速探索构象空间的局部区域。我们将该方案与两种广泛使用的模拟方法——复制交换分子动力学和多轨迹采样——进行了比较，以评估其在构建 Aβ构象文库方面的效率和可靠性。我们构建了每个文库中各种结构和能量特征的概率分布函数，并比较了这些方案中的多样性和收敛速度。研究结果表明，所提出的方案是一种成功的计算方法，可用于生成与实验数据一致的 Aβ单体的多样化构象文库。该方法侧重于在更短的计算时间内访问更多的构象，而不关注文库中每个状态的统计权重。我们相信，所提出的计算技术可用于生成合理的起始池，随后可以与实验数据进行重新加权，以获得统计集合。
