[gmx-users] A bug of trjconv for force output?

[LuLanyuan]

Hello All,
I just found a strange problem for the gromacs tool trjconv. I have a trajectory (and the tpr file) which contains three groups of molecules and x,v and f. When I use trjconv to output the trajectory of the 1st group, everything is fine. For example, I have A,B and C groups in my trajectory and I choose A only at the prompt of trjconv and use "-force" option to include forces in my output trajectory of A only. But if I want a trajectory of C only, I found the output forces were wrong. Positions and velocities are right in this case. It seems to me the trjconv just pick the first N force numbers to write where N is the total number of atoms in group C. In fact these forces belong to group A and B. This "logic" of trjconv explains why the output is correct if I choose the first group A to write.
Some month ago I reported a problem regarding the deshuf.ndx and the wrong force output. I think that problem is related to this problem. Because fewer people use forces, I think this is the reason these bugs haven't been found before. I appreciate if some authors can fix them.
Thanks.
Lanyuan Lu 
_________________________________________________________________
用手机MSN聊天写邮件看空间，无限沟通，分享精彩！
http://mobile.msn.com.cn/