AMBER Archive (2006)

Subject: Re: AMBER: how to include EXTRA POINTS in AMBER8

From: David A. Case (case_at_scripps.edu)
Date: Thu Dec 07 2006 - 00:14:44 CST

On Mon, Dec 04, 2006, Rachel wrote:
>
> Thanks for your reply. I have tried to put a finite mass on that extra point
> (I used mass = 1.008 as the normal hydrogen atom), and the result is even I
> used very very small timestep (e.g. 0.1fs), I don't have any problem with
> vlimit, and the error message is:
> ####################################################################
> NSTEP = 0 TIME(PS) = 0.000 TEMP(K) = 0.00 PRESS =
> 0.0
> Etot = -217128.3786 EKtot = 0.0000 EPtot = -
> 217128.3786
> BOND = 400.3220 ANGLE = 2165.2216 DIHED =
> 6798.3874
> 1-4 NB = 2513.3831 1-4 EEL = 34285.2873 VDWAALS =
> 25111.1035
> EELEC = -288402.0835 EHBOND = 0.0000 RESTRAINT =
> 0.0000
> Ewald error estimate: 0.8763E-04

Does minimization work, and give the same energy? If you are setting ntc>1,
the system may be trying to shake bonds to hydrogen atoms, and not know how to
handle your hydrogen molecules. I still think that you (or someone) is
probably going to have to really debug the code here, since you are trying
something that is well outside the ordinary design parameters of the code.

Try without shake; try making a very small system (peptide plus a couple of
H2's?), run short simulations, print every step, etc. See what happens
without charges on the hydrogen points. Basically, try to experiment to
narrow down the source of the problem. Then you can go back to your original
system.

...good luck...dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu