关于使用Martini力场模拟后计算扩散系数的问题

xiaojiao228

各位老师好，在使用Martini力场模拟完成后，通过获得轨迹的MSD曲线然后计算斜率得到组分的D值，在计算过程中，突然想到在原始文献中关于时间尺度的问题，不太确定，是不是需要在自己计算出来的D值得基础上乘上系数4得到的才是想要的那个扩散系数呢？原文如下：

Interpretation of Time Scale. In general, the interpreta-tion of the time scale in CG simulations is not straightforward.In comparison to atomistic models, the dynamics observed withCG models is faster. The main reason is that the underlyingenergy landscape is much smoother as a result of the largerparticle sizes. The friction arising from the atomic degrees offreedom is simply missing. On the basis of comparison ofdiffusion constants in CG and atomistically modeled systems,the effective time sampled using our CG model was found tobe 2- to 10-fold larger. When interpreting the simulation resultswith the CG model, one can to a first approximation simplyscale the time axis. Somewhat surprisingly, a similar scalingfactor appears to describe the general dynamics present in avariety of systems quite well. For instance, water permeationrates across a DPPC membrane and lipid lateral diffusion ratesare observed to be in good agreement with experimentalmeasurements after scaling the rates by a factor of 4.23 Moreglobal events, such as the aggregation of lipids into vesicles,also occur on a time scale comparable to atomistic simulationsafter 4-fold scaling.51 Likewise, sampling of the configurationalspace of liquid hydrocarbons and of the lipid tails inside a bilayerwas 5- to 10-fold faster for the CG model in comparison to anatomistic model46,47 if no scaling factor was applied. Thesimulation times reported in the remainder of the paper areeffective times, unless explicitly stated (denoted as simulationtime). The standard conversion factor we use is a factor of 4,which is the speed up factor in the diffusional dynamics of CGwater compared to real water.23 Note that the dynamics of thesystem is affected by the assignment of masses to the CG beads.For reasons of computational efficiency, standard masses m )72 amu (corresponding to four water molecules) are typicallyused. Especially for ring systems, however, assignment ofstandard masses to all sites will lead to significantly underes-timated dynamics. For accurate analysis of kinetic properties,more realistic masses should be used in the simulations.However, in homogeneous systems scaling of the kineticproperties based on mass differences can be performed after-ward.