Chambers Eric J, Price Eric A, Bayramyan Melina C, Haworth Ian S
Department of Pharmaceutical Sciences, University of Southern CaliforniaLos Angeles, CA 90089-9121, USA.
J Biomol Struct Dyn. 2003 Aug;21(1):111-25. doi: 10.1080/07391102.2003.10506909.
We present an algorithm for the computation of 2'-deoxyribose-phosphodiester backbone conformations that are stereochemically compatible with a given arrangement of nucleic acid bases in a DNA structure. The algorithm involves the sequential computation of 2'-deoxyribose and phosphodiester conformers (collectively referred to as a backbone 'segment'), beginning at the 5'-end of a DNA strand. Computation of the possible segment conformations is achieved by the initial creation of a fragment library, with each fragment representing a set of bond lengths, bond angles and torsion angles. Following exhaustive searching of sugar conformations, each segment conformation is reduced to a single vector, defined by a specific distance, angle and torsion angle, that allows calculation of the O(1)' position. A given 'allowed' conformation of a backbone segment is determined based on its compatibility with the base positions and with the position of the preceding backbone segment. Initial computation of allowable segment conformations of a strand is followed by the determination of continuous backbone solutions for the strand, beginning at the 3'-end. The algorithm is also able to detect repeating segment conformations that arise in structures containing geometrically repeating dinucleotide steps. To illustrate the utility and properties of the algorithm, we have applied it to a series of experimental DNA structures. Regardless of the conformational complexity of these structures, we are able to compute backbone conformations for each structure. Hence, the algorithm, which is currently implemented within a new computer program NASDAC (Nucleic Acids: Structure, Dynamics and Conformation), should have generally applicability to the computation of DNA structures.
我们提出了一种用于计算2'-脱氧核糖 - 磷酸二酯主链构象的算法，该构象在立体化学上与DNA结构中核酸碱基的给定排列兼容。该算法涉及从DNA链的5'-末端开始依次计算2'-脱氧核糖和磷酸二酯构象异构体（统称为主链“片段”）。通过最初创建片段库来实现对可能的片段构象的计算，每个片段代表一组键长、键角和扭转角。在对糖构象进行详尽搜索之后，每个片段构象都被简化为一个单一向量，该向量由特定的距离、角度和扭转角定义，从而可以计算O(1)'位置。主链片段的给定“允许”构象是根据其与碱基位置以及前一个主链片段位置的兼容性来确定的。在对链的允许片段构象进行初始计算之后，从3'-末端开始确定链的连续主链解决方案。该算法还能够检测在包含几何重复二核苷酸步的结构中出现的重复片段构象。为了说明该算法的实用性和特性，我们已将其应用于一系列实验性DNA结构。无论这些结构的构象复杂性如何，我们都能够计算每个结构的主链构象。因此，目前在新计算机程序NASDAC（核酸：结构、动力学和构象）中实现的该算法，应该对DNA结构的计算具有普遍适用性。
