Hi, Thomas
I modified thermo_int.f, in the way you told me, ran a shorter
simulation and here is a part of the output
DV/DL, AVERAGES OVER 50000 STEPS
NSTEP = 50000 TIME(PS) = 120.000 TEMP(K) = -0.17 PRESS = 0.0
Etot = -8.8950 EKtot = -8.8950 EPtot = -21.7097
BOND = 0.0000 ANGLE = 0.0000 DIHED = 0.0000
1-4 NB = 0.0000 1-4 EEL = 0.0000 VDWAALS = 9.3588
EELEC = 0.0000 EHBOND = 0.0000 RESTRAINT = -31.0684
EAMBER (non-restraint) = 9.3588
DV/DL = 9.3588
Now VDWAALS energy is equal to DV/DL and I'm satisfied :)
I recall that you told me that I was out for some testing, so, good
thing that I checked the results carefully.
You better put together a bug fix.
I also suggest that it will be more carefully written in the manual
that the restraints should be applied to V0 (and hence that V0 should
contain the uncharged ligand). I'm not sure that the code, written in
this way, will work for the other situation.
Best regards and thanks for a quick replay,
Samuel
2010/2/24 <steinbrt.rci.rutgers.edu>:
> Hi,
>
> what you write is correct. It appears to me on first glance that you have
> encountered a bug in applying dvdl_norest. In thermo_int.f the restraint
> energy on process V0 is subtracted from dvdl. I guess since we want to
> subtract the restraint energy from V0, which enters dvdl as V1-V0, that
> should be a plus instead. Embarrasing error :-(
>
> Try and change line 27 of thermo_int.f from
>
> dvdl = dvdl - ecopy(32)
>
> to
>
> dvdl = dvdl + ecopy(32)
>
> recompile and rerun a short simulation. You should get a dvdl that is
> indeed equal to VDWAALS in your case.
>
> The reason this remained undetected is that few people used the
> functionality and the restraint contribution is small enough so that in my
> test cases, I still got reasonable results as compared to experiment... If
> that turns out all right, I will put together a bug fix.
>
> Thanks for noticing (and apparently being the only one using that part of
> the code),
>
> Thomas
>
> Dr. Thomas Steinbrecher
> BioMaps Institute
> Rutgers University
> 610 Taylor Rd.
> Piscataway, NJ 08854
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 24 2010 - 11:00:02 PST
