Development of a System of Simulation Biopolymer Systems in Dihedral Angle Space

在过去的十多年中，我们一直在开发一个计算系统和软件系统，以实现二面角空间中生物聚合物体系构象模拟的计算。该项目的目的是（a）扩大这一计算的范围和（B）为向全世界用户提供这一计算做好准备。为了达到这些目的，我们取得了下列成果。(1)软件系统根据算法本身所包含的内在逻辑进行了重新设计。该系统具有通用性和灵活性。一篇描述这种重新表述的论文正在《计算机物理通讯》上出版。(2)二面角空间中的计算系统已被证明是非常成功的，适用于蛋白质。然而，五元核糖和脱氧核糖环的柔性阻碍了核酸的成功应用。为了克服这个困难，必须发展一种分析理论来处理核糖和脱氧核糖环的假旋转。我们已经成功地从根本上解决了这个问题。描述这一理论的一篇论文已发表在J.Phys.Chem.（3）我们已经发展了一个处理核酸的计算系统，它是通过结合上述处理五元环的伪旋转的方法来实现的。更具体地说，我们已经开发了一种算法，从实验NMR信息获得的距离信息推断核酸的三维结构。该算法允许精确地处理核糖和脱氧核糖环的柔性。该算法不仅速度快，而且比现在流行的笛卡尔坐标空间中的算法更少的偏见。

During the last ten or more years we have been developing a system of computation and a software system to implement the computation for simulation of conformations of biopolymer systems in the dihedral angle space. The purpose of this project is (a) to expand the scope of this computation and (b) to prepare for making it available for world-wide users. As to these purposes we have attained the following results. (1) The system of software has been reformulated according to the intrinsic logic which the algorithm contains inherently. Accordingly the system attained a generality and flexibility. A paper describing this reformulation is in press in Computer Physics Communication. (2) The system of computation in the dihedral angle space has proved very successful as applied for proteins. However, successful application for nucleic acids has been hampered by the flexibilities of the five-memebered ribose and deoxyribose rings. In order to overcome this difficulty, an analytic theory to treat pseudo-rotation of ribose and deoxyribose rings has to be developed. We have succeeded to give a fundamental solution to this problem. A peper describing this theory has been published in J.Phys.Chem. (3) We have developed a system of computation for treating nucleic acids by incorporating the above method of treating the pseudo-rotation of the five-membered rings. To be more specific, we have developed an algorithm to deduce three dimensional strucutres of nucleic acids from distance information obtainable from experimental NMR information. This algorithm allows to treat accurately the flexibities of ribose and deoxyribose rings. This algorithm is not only faster but also less biased than the now popular algorithm that works in the Cartesian coordinate space.