AMBER Archive (2004)

Subject: AMBER: GB and Atom Type

From: Steve Seibold (seibold_at_chemistry.msu.edu)
Date: Tue Jan 20 2004 - 09:34:45 CST

Dear Amber Users

I am trying to do MD using implicit solvent models on a protein that
contains a heme iron.

Using, in-vacuo, 8.0 angstrom cutoff, the distance dependent dielectric,
the MD works fine.

However, once I switched to the Generalized Born solvation , using the
same protein topology file used in the distance dependent run in vacuo
above, the simulation failed to run and gave the following message: "bad
atom type: FE".

I used the "check" command in xleap to check the unit "HEM" and found
that although there are 3 "close contacts" between certain "carbon
atoms" overall the unit was "ok".

I do have a total charge of -1.00 for the heme.

Is there some parameter in the input file (e.g. CUT ?) that I have to
experiment with due to the iron? Do I need to change the iron parameters
for GB?

The input file I have used is:

Protein w/ 30.0 GB, constant Volume

&cntrl

ntx = 1, irest = 0, ntrx = 1, ntxo = 1,

ntpr = 5000, ntwx = 10000, ntwv = 0, ntwe = 0,

ntwr = -5000,

ntf = 2, ntb = 0,

cut = 30.0, nsnb = 10,

ibelly = 0, ntr = 0,

igb = 1, saltcon = 0.2, gbsa = 1,

imin = 0,

nstlim = 400000,

nscm = 1000,

t = 0.0, dt = 0.002,

temp0 = 300.0, tempi = 100.0,

ig = 71277, heat = 0.0,

ntt = 1,

tautp = 0.2,

vlimit = 15.0,

ntp = 0, pres0 = 1.0, comp = 44.6,

taup = 0.2, npscal = 1,

ntc = 2, tol = 0.00001,

lastrst = 8000000,

&end

END

I would appreciate any assistance anyone could offer.
 
Thanks, Steve

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