Guvench Olgun, Brooks Charles L
Department of Molecular Biology (TPC-6), The Scripps Research Institute, 10550 North Torrey Pines Rd., La Jolla, California 92037, USA.
J Comput Chem. 2004 Jun;25(8):1005-14. doi: 10.1002/jcc.20026.
Continuing advances in computer hardware and software are permitting atomic-resolution molecular simulations for longer time scales and on larger systems. Despite these advances, routinely performing atomistic simulations with explicit water for even small proteins, which reach the folding time of such proteins, remains intractable for the foreseeable future. An implicit approximation of the solvent environment using a solvent accessible surface area (SASA) term in a molecular mechanics potential function allows exclusion of the explicit water molecules in protein simulations. This reduces the number of particles by approximately an order of magnitude. We present a fast and acceptably accurate approximate all-atom SASA method parameterized using a set of folded and heat-denatured conformations of globular proteins. The parameters are shown to be transferable to folded and heat-denatured conformations for another set of proteins. Calculation of the approximate SASA and the associated derivatives with respect to atomic positions for a 4644 atom protein requires only 1/11th the CPU time required for calculation of the nonbonded interactions for this system. On a per atom basis, this algorithm is three times faster than the fastest previously published approximate SASA method and achieves the same level of accuracy.
计算机硬件和软件的不断进步使得在更长的时间尺度和更大的系统上进行原子分辨率的分子模拟成为可能。尽管有这些进展，但在可预见的未来，即使对于小蛋白质，要使用显式水进行常规的原子模拟以达到此类蛋白质的折叠时间，仍然是难以处理的。在分子力学势函数中使用溶剂可及表面积（SASA）项对溶剂环境进行隐式近似，可以在蛋白质模拟中排除显式水分子。这将粒子数量减少了大约一个数量级。我们提出了一种快速且精度可接受的近似全原子SASA方法，该方法使用一组球状蛋白质的折叠和热变性构象进行参数化。结果表明，这些参数可转移到另一组蛋白质的折叠和热变性构象上。对于一个4644个原子的蛋白质，计算近似SASA及其相对于原子位置的相关导数仅需要该系统计算非键相互作用所需CPU时间的1/11。按每个原子计算，该算法比之前发表的最快的近似SASA方法快三倍，并且达到相同的精度水平。
