Schellman J A, Harvey S C
Institute for Molecular Biology, University of Oregon, Eugene, USA.
Biophys Chem. 1995 Jun-Jul;55(1-2):95-114. doi: 10.1016/0301-4622(94)00144-9.
Long molecules of DNA have the statistical properties of a worm-like coil. Deviations from linearity occur both because of small dynamic bends induced by thermal motion and from a random distribution of static bends. The latter originate in the different conformations of each of the possible base pair sequences. In this paper a statistical theory of the persistence length of DNA is developed which includes both static and dynamic effects for each base pair sequence, as well as the sequence-dependent correlations of bending angles. The result applies to a generic DNA, i.e., the average over an ensemble of all possible sequences. The theory is also applied to the generation of the average properties of curved DNAs by an analytic method that includes dynamic averaging as well as correlated bends. These results provide information which supplements that obtained by others using Monte Carlo methods. The additivity relation 1/P = 1/P(S) + 1/P(d) proposed by Trifonov et al., where P is the persistence length and P(S) and P(d) are the persistence lengths arising from purely static and dynamic effects, respectively, has been verified to be accurate to better than 0.5%. This is true for both a simplified model and one that includes a complete set of static bends at all base pair sequences.
DNA的长分子具有类蠕虫状螺旋的统计特性。由于热运动引起的小动态弯曲以及静态弯曲的随机分布，都会导致分子偏离线性。后者源于每个可能的碱基对序列的不同构象。本文建立了一种DNA持久长度的统计理论，该理论涵盖了每个碱基对序列的静态和动态效应，以及弯曲角度的序列依赖性相关性。结果适用于一般的DNA，即所有可能序列集合的平均值。该理论还通过一种解析方法应用于弯曲DNA平均性质的生成，该方法包括动态平均以及相关弯曲。这些结果提供的信息补充了其他人使用蒙特卡罗方法获得的信息。Trifonov等人提出的加和关系1/P = 1/P(S) + 1/P(d)，其中P是持久长度，P(S)和P(d)分别是由纯静态和动态效应产生的持久长度，已被验证其准确性优于0.5%。对于简化模型和包含所有碱基对序列完整静态弯曲集的模型都是如此。
