Hi all,
**********************************************************
This is the second try to get your reply.
Please take a look at this problem and give me a word.
**********************************************************
I'm doing md simulations of a small peptide in water and one of amino acids
of the peptide is phosphoserine.
The atomic partial charges of singly and doubly charged phosphoserines were
calculated using antechamber.
Both trajectories were run at 500 K using the parm99 parameters.
But the problem is that the trajectory of the peptide having singly charged phosphoserine
failed to run after a few picoseconds with vlimit exceeded while heated from 0 K to 500 K.
Structures of the peptide were severely distorted and looked unreasonable.
On the other hand, there was no problem in the case having doubly charged phosphoserine.
Several other test simulations were done using a smaller time step, without using SHAKE for hydrogens, or by heating stepwisely, but still the trajectory failed with vlimit exceeded.
So, I checked the parameters and found that the force constant for the angle O2-P-OH is
only about half of the value in CHARMM.
O2-P-OH 45.0 108.23 in Amber/parm99.dat
O2-P-OH 98.9 108.23 in CHARMM/pardna_10_22.inp
(ON3-P-ON4 98.9 108.23 in CHARMM/par_all27_prot_na.prm)
In a new test simulation using the force constant larger than 45.0 for the angle O2-P-OH,
for example 60.0, there was no problem in keeping the simulation running without vlimit exceeded.
I'm wondering that the problem in simulating the peptide having singly charged phosphoserine
lies in using the rather small force constant or in anything else.
For reference, the input parameters used in heating the system are:
&cntrl
imin = 0, ntx = 1, irest = 0,
ntpr = 500, ntave = 5000,
ntwr = 500, iwrap = 0,
ntwx = 0, ntwe = 0,
ntf = 1, ntc = 2, ntb = 1,
cut = 8.0, nsnb = 10,
nstlim = 50000, dt = 0.002,
tempi = 0.0, temp0 = 500.0,
ntt = 1, tautp = 1.0
&end
Thanks for your attention and please give me some nice idea to get out of this problem.
Seongeun Yang
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Thu Aug 12 2004 - 14:53:01 PDT
