Hi Jordan,
> I am trying to look at the water on a model but the H2O molecules
> looks pretty weird... You will find a pdb as attachment to have a
> look, you can notice that the H2O molecules have some odd positions/
> contacts around DNA. My question is : where does this problem come
> from? Anyway, it does not seem to interfere on the global energy which
> looks correct
This does look pretty weird. Do you have the prmtop and original inpcrd
file? Plus the restart file produced by the run? This looks to me like there
must be some issue with the initial structure, or something is going bad but
it isn't obvious from the energies. What does the initial structure look
like. If you watch the trajectory file can you see where things start going
weird?
I would also ask - did all the test cases pass?
A few comments on your input file...
> Here is my sander input file :
>
> &cntrl
> imin = 0, ntx = 7, irest = 1,
> timlim = 999999, ig = 99775, ibelly = 0, ntr = 0,
timlim I do not think exists anymore in AMBER 10 and I have no idea what it
does in this context.
> ntb = 2,
> dielc = 1, cut = 9.0, scnb = 2.0, scee = 1.2,
> ntc = 2, tol = 0.0005, ntf = 2,
0.0005 is a VERY loose tolerance for the shake algorithm and could be the
cause of your problems. I would suggest leaving this at the default or even
making it an order of magnitude tighter (10^-6).
> tempi = 300.0, temp0 = 300.0,
> ntt = 1, tautp = 0.2,
> ntp = 1, pres0 = 1.0, taup = 0.2,
These are very quick coupling constants for tautp and taup you might want to
consider reducing them to 1.0 or larger. Especially taup.
> nstlim = 1000000, dt = 0.002, ndfmin = 6, ntcm = 1, nscm = 100,
ndfmin and ntcm are 'very' deprecated. I have to go back to AMBER v7.0 to
even find a description in the manual. They just get ignored in AMBER 10 I
think but you should probably remove them from your input file to be sure.
Also nscm of 100 is probably a little too frequently to be removing COM.
> trajin model19.mdcrd
> center :1-4
> image familiar
> trajout traj.pdb pdb
> go
Looks fine to me but I would be interested in what the actual 'raw' mdcrd
file looks like since this will give you a hint as to whether it is a sander
problem or a ptraj problem.
All the best
Ross
/\
\/
|\oss Walker
| Assistant Research Professor |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
|
http://www.rosswalker.co.uk | PGP Key available on request |
Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Jun 09 2009 - 01:16:06 PDT
