AMBER Archive (2005)

Subject: RE: AMBER: Extended bonds-MD

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Tue Jun 14 2005 - 12:28:43 CDT


Dear Ananda

A few comments on your input options:

> MD_simulation of IMB in PBC_1
> &cntrl
> imin = 0, irest = 0, ntx = 1,
> ntt = 1, temp0 = 10.0,

temp0=10.0 you only want a temperature of 10 kelvin correct?

> tautp = 1, gamma_ln = 10000,

gamma_ln is not used when ntt=1 only with ntt=3. For the record though a
value of 10000 for gamma_ln is huge. A value of 4 or 5 would be more
appropriate.

> ntp = 0,
> ntb = 1, ntc = 1, ntf = 7,

ntf=7 -> This almost certainly the cause of your problems. This means omit
all bond angle and dihedral interactions. Do you really want to do this? In
my experience running a simulation of a protein with just non-bonded
interactions is never stable. The only time I find you can get away with
NTF>1 is for NTF=2 and only then when you have shake turned on (ntc=2).
Also, since you are running a 0.5fs time step there is nothing to really be
gained from ommiting bonds angles and dihedrals. I would treat ntf values
greater than 2 as being developer/debugging options.

> cut = 20.0,

20 angstroms is probably much much bigger than you need for a periodic
boundary simulation. Set it to 8 angstroms and only increase it if you don't
run with PME.

>scee = 4.0, scnb = 20.0,

Why have you changed these? These really are developer options and should
"never?" be changed by the end user without good reason. The entire force
field (FF99, FF03...) is parameterised around the default values of these
scaling factors. SCNB should be 2.0 and scee should be 1.2. Only change them
if you have for some reason built yourself a completely new force field.

> keep PIB fixed with strong constraints
> 100
> RES 1 64
> END
> END

100.0 KCal/mol is a pretty big restraint and may cause instability in your
system. It is probably ok for minimisation but I would reduce this during
molecular dynamics. A value of 10 or less should be sufficient to keep these
residues fixed reasonably close to the initial structure.

> is it due to the less simulation time period or the effect of box size
> or the effect of some other parameters.

This is an artifact of the various "strange" options in your input file that
I have highlighted above.

All the best
Ross

/\
\/
|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel:- +1 858 784 8889 | EMail:- ross_at_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.

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