Suri A K, Levy R M
Department of Chemistry, Rutgers University, Piscataway, New Jersey 08855-0939.
J Magn Reson B. 1995 Jan;106(1):24-31. doi: 10.1006/jmrb.1995.1004.
Long-mixing-time data (tau m > 200 ms) from NOE spectra have largely been ignored as a source of protein structural information due to the effects of spin diffusion on calculated interproton distances when using the two-spin approximation. An effective approach for incorporating spin-diffusion effects in an average way into refinements is to choose distance bounds based on distributions of distances observed in NOE back calculations on homologous proteins from a protein structure database. We have determined distributions of interproton distances characteristic of newly observed NOE cross peaks for the proteins crambin, PTI, and echistatin at long mixing times. A relaxation-matrix analysis was used to model the effects of spin diffusion. Constraint ranges were constructed from the interproton distance distributions which can be used in standard protein-refinement programs based on the two-spin approximation. Back calculations are also used to analyze constraint ranges typically used for protein structure determinations based on NOE spectra at shorter mixing times.
由于在使用双自旋近似时自旋扩散对计算出的质子间距离有影响，来自NOE谱的长混合时间数据（τm > 200 ms）在很大程度上被忽视，未能作为蛋白质结构信息的来源。一种将自旋扩散效应以平均方式纳入精修的有效方法是，根据在蛋白质结构数据库中同源蛋白质的NOE反向计算中观察到的距离分布来选择距离界限。我们已经确定了在长混合时间下，蛋白质胰凝乳蛋白酶原、蛋白酶抑制剂I（PTI）和蛇毒抑肽中新观察到的NOE交叉峰所特有的质子间距离分布。采用弛豫矩阵分析来模拟自旋扩散的影响。基于质子间距离分布构建了约束范围，这些范围可用于基于双自旋近似的标准蛋白质精修程序。反向计算还用于分析通常用于基于较短混合时间的NOE谱进行蛋白质结构测定的约束范围。
