Department of Nuclear Physics, University of Madras, Chennai, Tamil Nadu, India.
Centre of Advanced Study in Crystallography and Biophysics, University of Madras, Chennai, Tamil Nadu, India.
Phys Chem Chem Phys. 2024 Aug 28;26(34):22640-22655. doi: 10.1039/d4cp01891e.
We propose digital nets conformational sampling (DNCS) - an enhanced sampling technique to explore the conformational ensembles of peptides, especially intrinsically disordered peptides (IDPs). The DNCS algorithm relies on generating history-dependent samples of dihedral variables using bitwise XOR operations and binary angle measurements (BAM). The algorithm was initially studied using met-enkephalin, a highly elusive neuropeptide. The DNCS method predicted near-native structures and the energy landscape of met-enkephalin was observed to be in direct correlation with earlier studies on the neuropeptide. Clustering analysis revealed that there are only 24 low-lying conformations of the molecule. The DNCS method has then been tested for predicting optimal conformations of 42 oligopeptides of length varying from 3 to 8 residues. The closest-to-native structures of 86% of cases are near-native and 24% of them have a root mean square deviation of less than 1.00 Å with respect to their crystal structures. The results obtained reveal that the DNCS method performs well, that too in less computational time.
我们提出了数字网络构象采样（DNCS） - 一种增强的采样技术，用于探索肽，特别是天然无序肽（IDP）的构象集合。DNCS 算法依赖于使用按位异或操作和二进制角度测量（BAM）生成依赖于历史的二面角变量样本。该算法最初使用高度难以捉摸的神经肽 met-enkephalin 进行了研究。DNCS 方法预测了接近天然的结构，并且观察到 met-enkephalin 的能量景观与先前对神经肽的研究直接相关。聚类分析表明，分子只有 24 种低能构象。然后，已经使用 DNCS 方法来预测长度从 3 到 8 个残基的 42 个寡肽的最佳构象。86％的情况下最接近天然的结构是近天然的，其中 24％的结构与晶体结构的均方根偏差小于 1.00Å。所获得的结果表明，DNCS 方法表现良好，而且计算时间也较短。
