[gmx-users] trajectory info

[Mark Abraham]

D D wrote:
> Hi *,
> I wrote a tool to calculate properties from trajectories (parallel 
> 4.0.5), but it seems that I miss something.
> In the case of MSD, I do not get the curve that g_msd produces for mol com.
> On the other hand, if I run this tool for MSD on a trajectory produced 
> by trjconv -pbc mol,
> I get the right curve (I handle pbc jumps of com after calc_xcm).
> I think that the problem comes from pbc handling for molecules splited 
> across boundaries - I expect : do something before calc_xcm.
> I could run trjconv before run tool, but I 'd like to use the original 
> trajectories and handle them correctly.
> Any suggestion is appreciable.

Look at the routine trjconv -pbc mol is using.

Mark

> The procedure I follow is :
> read_tps_conf
> read_first_frame
> do {
> prepare_index_etc
> calc_xcm
> do_some_calculations_here
> }while (read_next_frame)
> 
> Thanks
> DD
>