negative grid value for z-path
Ben Cossins
I have been testing plumed for protein conformational change
calculations using Gromacs. I applied all the extra patches you have
posted.
I have been using the s-path and z-path RMSD CV's with 20 frames and
70 atoms being aligned and displaced with weighting 1.00 in every
case. The msd between frames is around 0.15 so I have been using
quite a high lambda of about 18.
If I try and use GRID (min 0, max 20) the run dies at 0 ps saying
it is outside of the GRID. I have had a look in the code and the
value of the grid for the zpath is -0.09. If I set the grid min to
0.5 then the calculation seems to run fine.
Could you help me interpret these issues?
Many thanks
Ben
davideb
You don't need to be worried by negative Z values. Your value for
lambda seems correct
(although, given your average MSD I would have used 15 instead) and
the negative value
comes from the fact that actually you're calculating a negative log
for a value which is
larger than 1 (because you may sit on one frame and have distance zero
from it but the neighbor
frames may contribute as well giving a sum of exponentials which is
larger than one).
This sometimes happens (I also find rather unpleasant but I decided I
can live with it).
Concerning the GRID usage I'm not an expert so I prefer to leave this
topic to some of the developers who
coded it!
Good luck!
D
PS: it results you have not subscribed the list yet. It's our policy
to answer only to subscribers. Can you make it ?
Massimiliano Bonomi
> If I try and use GRID (min 0, max 20) the run dies at 0 ps saying
> it is outside of the GRID. I have had a look in the code and the
> value of the grid for the zpath is -0.09.
Dear Ben,
the interval min 0 and max 20 should be fine for the s-path variable (I would use 0 and 21 just to be sure...).
Which boundaries did you use for the z-path variable?
Max
> If I set the grid min to
> 0.5 then the calculation seems to run fine.
> Could you help me interpret these issues?
>
> Many thanks
>
> Ben
>
> --
> You received this message because you are subscribed to the Google Groups "PLUMED users" group.
> To post to this group, send email to plumed...@googlegroups.com.
> To unsubscribe from this group, send email to plumed-users...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/plumed-users?hl=en.
>
Ben Cossins
I would be happy to subscribe. How do I go about doing that. I
have become a member of the google group, at least I thought I had.
Thanks
ben
Ben Cossins
Initially I used min 0 and max 20 for the z-path as I hadn't been
through the z-path equation properly. Now I am using min -1.0 max 1.0
and with 50 bins. The z-path value seems to stay around -0.09 so far
and fluctuates no more that 0.02.
Thanks for your help
On Feb 22, 2:28 pm, Massimiliano Bonomi