Good evening,
I am having a problem with high velocities cropping up in my QM/MM
simulation. When done with the same protocols (input files, stages of
minimizing, heating) but purely MM (default "classical" force field
parameters for the old QM region) has no such issues. This would seem to
imply that there is an issue created at the boundary. I am not
(consciously) cutting across any bonds. This occurs after 20ps.
I ran a purely force-field based heating calculation before beginning
the QM/MM run. I am debating if it's possible that a discontinuity
emerges in going from the classical heating to the QM/MM run that
results in the overlapping atoms at the boundary (I have not yet been
able to observe, directly, overlapping atoms, merely inferring this from
the high velocity). I would think this unlikely; I would think the high
velocity would be seen immediately, not after 20ps of running.
When looking through the Amber archives, I specifically find the
opposite: people report doing their heating classically before QM/MM
just fine (and is alleged to be more logical).
My question: does blaming a non QM/MM heating seem at all plausible if
the problem emerges after 20ps?
I am running the calculation out of desperation to test it, but I think
it will fail (in 2 days time...). I have attached my input file for the
production NPT run, in case I am doing something obviously stupid
unknowingly.
Production
&cntrl
imin=0,
ntx=5,
irest=1,
nstlim=500000,
dt=0.002,
ntf=2,
ntc=2,
temp0=300.0,
ntpr=100,
ifqnt=1
ntwx=100,
cut=10.0,
ntb=2,
ntp=1,
ntt=3,
gamma_ln=1.0,
ig=-1,
/
&qmmm
iqmatoms=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,
qmcut=10.0,
qm_pme=0
qm_ewald=0
qmshake=0,
qm_theory = 'EXTERN',
writepdb=1,
/
&gau
method = 'm062x',
basis = '6-31G**',
num_threads = 16,
scf_conv= 6,
mem='10GB',
/
--
Dr. Robert Molt Jr.
r.molt.chemical.physics.gmail.com
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 20 2015 - 21:00:03 PDT
