Subject: Re: AMBER: GIBBS - PMF problem: Lambda has not changed
From: David Smith
Date: 14 Nov 2003 11:50:01 +0100
Hello again Jiri,
I'll try to offer a few constructive comments:
If you use ntx=7 you probably also want to set init=4 to get your
starting velocities from the restart.
I would probably write a trajectory a little more often in this example
(eg ntwx at most 100 so that you get at least one snapshot per window).
Similarly I would probably have ntpr =< 100 to get a bit more
information during the run.
I would only use itimth=1 if you absolutely have to (i.e. the bond is
part of a closed ring). This method seems to converge significantly more
slowly than the PF method. It also has a complication about the moment
of inertia.
Other than that, I can't for the life of me figure out what's wrong.
I ran the following (slighlty modified) version of your input using my
own topology, coordinates, and constraints:
********************
test for Jiri
&cntrl
ntx=6, (I happened to have a binary restart)
ntb=2,
ntp=1,
ntwx=250,
ntwe=250,
isande=1,
ntpr=1,
scee=1.2,
cut=15.0,
temp0=300,
taup=1.0,
dt=0.002,
ntc=2,
ntf=2,
ntt=1,
tautp=0.2,
nscm=-1,
nrun=2, (to make it faster)
nstlim=10,
isldyn=3,
idifrg=1,
almda=0.8, (this happened to be the lamba from my rst)
almdel=0.01, (I only wanted to step a little)
nstmeq=3, (faster)
nstmul=7, (ditto)
ncorc=1,
intr=1,
itimth=1,
&end
06595 11954 00000 00000 00000 00000 00002 000.00000 001.00000
0000.00000 008.00000 000.00000 038.41020 0000 0000
*****************************
and everything worked fine:
test for Jiri
/FREE ENERGIES/
FORWARD # STEPS EQUILIBRATION DATA
COLLECTION
WIND LAMBDA dLAMBDA NR EQ COLL FORWARD REVERSE FORWARD
REVERSE
1 0.800000 0.010000 0 3 7
2 0.810000 0.010000 0 3 7 5.7404 1.0940
*SUMMARY*
FREE ENERGIES
# STEPS DATA COLLECTION
AVE TOT AVE AVE TOTAL TOTAL
WIND STEPS LAMBDA dLAMBDA NR EQ COLL FORWARD REVERSE
2 20 0.80-0.81 0.010000 0 3 7 1.0940 -1.0940
I would check again to make sure your constraints are being read
properly. Also take a look at the section:
Constrained Internals at Last F.E. Update Lam L+d_L L-d_L
SG ( 6595)-S48 ( 11954) 32.6323 32.9364 32.3282
to see what it says.
If still nothing than I can only suggest that you pack up for topology,
restart, input and script into an archive (prefereably tar.gz) and send
it my way. Probably best to avoid the list for that...
Good luck.
David.
On Thu, 2003-11-13 at 13:32, Jiri Matousek wrote:
> Hi Amber users,
> I am trying to do free energy calculation using PMF. The change in
> conformation depends on one distance, so I have one constraint. Problem
> is that lambda ( ... distance) is all the time at the starting value.
> I mostly used TI, but I tryied nearly all methods (by chaging isldyn)
> with the same results. No error message appered during my calculation,
> so I really have no suggestion what is wrong in my input file.
> Any answer will be very helpfull.
> Thank you in advance.
>
> Jiri Matousek
>
> Here is my input file and MICSTAT:
>
> input:
> #pokus
> &cntrl
> ntx=7,
> ntb=2,
> ntp=1,
> ntwx=250,
> ntwe=250,
> isande=1,
> ntpr=250,
> scee=1.2,
> cut=15.0,
> temp0=300,
> taup=1.0,
>
> dt=0.002,
>
> ntc=2,
> ntf=2,
>
> ntt=1,
> tautp=0.2,
> nscm=-1,
>
> nrun=11,
> nstlim=100,
>
> isldyn=3,
> idifrg=1,
> almda=0.0,
> almdel=0.1,
> nstmeq=30,
> nstmul=70,
>
> ncorc=1,
> intr=1,
> itimth=1,
> &end
> 14 41 0 0 0 0 2 0.00000 1.00000
> 100.00000 5.40000 100.00000 10.40000 0 0
>
>
>
> MICSTAT:
> #pokus
> /FREE ENERGIES/
> FORWARD # STEPS EQUILIBRATION DATA
> COLLECTION
> WIND LAMBDA dLAMBDA NR EQ COLL FORWARD REVERSE FORWARD
> REVERSE
> 1 0.000000 0.000000 0 30 70
> 2 0.000000 0.000000 0 30 70
> 3 0.000000 0.000000 0 30 70
> 4 0.000000 0.000000 0 30 70
> 5 0.000000 0.000000 0 30 70
> 6 0.000000 0.000000 0 30 70
> 7 0.000000 0.000000 0 30 70
> 8 0.000000 0.000000 0 30 70
> 9 0.000000 0.000000 0 30 70
> 10 0.000000 0.000000 0 30 70
> 11 0.000000 0.000000 0 30 70
>
> *SUMMARY*
> FREE ENERGIES
> # STEPS DATA COLLECTION
> AVE TOT AVE AVE TOTAL
> WIND STEPS LAMBDA dLAMBDA NR EQ COLL FORWARD
> 11 1100 0.00-0.00 0.000000 0 30 70 0.0000
>
>