Ming Dengming, Wall Michael E
Computer and Computational Sciences Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, USA.
Phys Rev Lett. 2005 Nov 4;95(19):198103. doi: 10.1103/PhysRevLett.95.198103. Epub 2005 Nov 2.
We propose a criterion for optimal parameter selection in coarse-grained models of proteins and develop a refined elastic network model (ENM) of bovine trypsinogen. The unimodal density-of-states distribution of the trypsinogen ENM disagrees with the bimodal distribution obtained from an all-atom model; however, the bimodal distribution is recovered by strengthening interactions between atoms that are backbone neighbors. We use the backbone-enhanced model to analyze allosteric mechanisms of trypsinogen and find relatively strong communication between the regulatory and active sites.
我们提出了一种蛋白质粗粒度模型中最优参数选择的标准,并开发了一种改进的牛胰蛋白酶原弹性网络模型(ENM)。胰蛋白酶原ENM的单峰态密度分布与从全原子模型获得的双峰分布不一致;然而,通过加强主链相邻原子之间的相互作用可以恢复双峰分布。我们使用主链增强模型来分析胰蛋白酶原的变构机制,并发现调节位点和活性位点之间存在相对较强的通讯。