Optimization of algorithms for ion mobility calculations
Optimization of algorithms for ion mobility calculations
复制标题
DOI：
10.1021/jp066953m
复制
发表时间：
2007-03-15
影响因子：
2.9
通讯作者：
Smith, Richard D.
中科院分区：
文献类型：
--
作者：
Shvartsburg, Alexandre A.;Mashkevich, Stefan V.;Smith, Richard D.
离子迁移谱（IMS）越来越多地用于探测气相离子的结构，特别是蛋白质和其他生物大分子的结构。这个过程涉及到比较测量的迁移率计算的潜在的几何形状，这需要使用一些近似处理的离子缓冲气体碰撞的取向平均截面的评价。两种常见的模型是投影近似（PA）和精确硬球散射（EHSS），将离子表示为硬球的集合。虽然大离子和/或构象集合的计算需要大量时间，但尚未探索算法优化。以前的EHSS程序主要由离子旋转操作，允许取向平均。我们已经开发了两种新的算法PA和EHSS计算：一个简化这些操作，大大减少了他们的数量，和其他处置他们完全从一个随机的起源传播轨迹。针对一组具有代表性的七种离子几何形状（包括不同的尺寸和形状）对新算法进行了测试。虽然最佳选择以一种不明显的方式取决于几何形状，但两种代码之间的差异通常是适度的。两者都比现有的软件更有效，例如比广泛使用的Mobcal（实现EHSS）快10-30倍。
Ion mobility spectrometry (IMS) is increasingly employed to probe the structures of gas-phase ions, particularly those of proteins and other biological macromolecules. This process involves comparing measured mobilities to those computed for potential geometries, which requires evaluation of orientationally averaged cross sections using some approximate treatment of ion-buffer gas collisions. Two common models are the projection approximation (PA) and exact hard-spheres scattering (EHSS) that represent ions as collections of hard spheres. Though calculations for large ions and/or conformer ensembles take significant time, no algorithmic optimization had been explored. Previous EHSS programs were dominated by ion rotation operations that allow orientational averaging. We have developed two new algorithms for PA and EHSS calculations: one simplifies those operations and greatly reduces their number, and the other disposes of them altogether by propagating trajectories from a random origin. The new algorithms were tested for a representative set of seven ion geometries including diverse sizes and shapes. While the best choice depends on the geometry in a nonobvious way, the difference between the two codes is generally modest. Both are much more efficient than the existing software, for example faster than the widely used Mobcal (implementing EHSS) similar to 10-30-fold.