Rutenber E, Katzin B J, Ernst S, Collins E J, Mlsna D, Ready M P, Robertus J D
Clayton Foundation Biochemical Institute, Department of Chemistry and Biochemistry, University of Texas, Austin 78712.
Proteins. 1991;10(3):240-50. doi: 10.1002/prot.340100308.
The plant cytotoxin ricin consists of two disulfide-linked chains, each of about 30,000 daltons. An initial model based on a 2.8 A MIR electron density map has been refined against 2.5 A data using rounds of hand rebuilding coupled with either a restrained least squares algorithm or molecular dynamics (XPLOR). The last model (9) has an R factor of 21.6% and RMS deviations from standard bond lengths and angles of 0.021 A and 4.67 degrees, respectively. Refinement required several peptide segments in the original model to be adjusted translationally along the electron density. A wide range of lesser changes were also made. The RMS deviation of backbone atoms between the original and model 9 was 1.89 A. Molecular dynamics proved to be a very powerful refinement tool. However, tests showed that it could not replace human intervention in making adjustments such as local translations of the peptide chain. The R factor is not a completely satisfactory indicator of refinement progress; difference Fouriers, when observed carefully, may be a better monitor.
植物细胞毒素蓖麻毒素由两条通过二硫键连接的链组成，每条链的分子量约为30,000道尔顿。基于2.8埃的多重同晶置换（MIR）电子密度图建立的初始模型，通过一轮轮的手工重建，并结合约束最小二乘法算法或分子动力学（XPLOR），根据2.5埃的数据进行了优化。最终模型（9）的R因子为21.6%，与标准键长和键角的均方根偏差分别为0.021埃和4.67度。优化过程需要对原始模型中的几个肽段沿电子密度进行平移调整。还进行了一系列较小的改动。原始模型与模型9之间主链原子的均方根偏差为1.89埃。分子动力学被证明是一种非常强大的优化工具。然而，测试表明它无法取代人工干预来进行诸如肽链局部平移等调整。R因子并不是优化进展的完全令人满意的指标；仔细观察的差分傅里叶图可能是更好的监测手段。
