On Tue, Sep 25, 2012 at 8:14 AM, Sangmin Lee <dadu0413.yonsei.ac.kr> wrote:
> Dear amber developers and users.
>
> I need to read scee and scnb from topology file in NAB. So I revised
> source files to set up pointers to store the variable 1-4 scaling arrays in
> the parm struct (AmberTools/src/sff/sff.h).
>
> > sff.h
>
> typedef struct parm {
> REAL_T *Gvdw, *Scee, *Scnb;
> }
>
> Also SCEE_SCALE_FACTOR and SCNB_SCALE_FACTOR set up in prm.c
> file(AmberTools/src/sff/).
>
> > prm.c
>
> /*
> * READ Scee -Scee
> */
> pfind(file, newparm, "SCEE_SCALE_FACTOR");
> for (i = 0; i Nptra; i++) {
> if (get_mytaskid() == 0) {
> #ifdef NAB_DOUBLE_PRECISION
> fscanf(file, " %lf", &prm->Scee[i]);
> #else
> fscanf(file, " %f", &prm->Scee[i]);
> #endif
> }
> }
> #if defined(MPI) || defined(SCALAPACK)
> MPI_Bcast(prm->Scee, prm->Nptra, MPI_DOUBLE, 0, MPI_COMM_WORLD);
> #endif
> skipeoln(file);
>
> /*
> * READ Scnb -Scnb
> */
> pfind(file, newparm, "SCNB_SCALE_FACTOR");
> for (i = 0; i Nptra; i++) {
> if (get_mytaskid() == 0) {
> #ifdef NAB_DOUBLE_PRECISION
> fscanf(file, " %lf", &prm->Scnb[i]);
> #else
> fscanf(file, " %f", &prm->Scnb[i]);
> #endif
> }
> }
> #if defined(MPI) || defined(SCALAPACK)
> MPI_Bcast(prm->Scnb, prm->Nptra, MPI_DOUBLE, 0, MPI_COMM_WORLD);
> #endif
> skipeoln(file);
>
>
> I need to know more details after above procedures.
>
> And I saw related mail by Jason in University of Florida. He said that I
> need to modify the 1-4 calculation in the nbond() routine of
> AmberTools/src/sff/sff2.c(There is only nbond2() routine. Is it right?) to
> use these values instead of scee (which is passed in as eelfac and enbfac).
> I would like to know more about it.
>
Yes, nbond2(), my mistake. As for needing to know more details, the only
way someone would know more details is if they had already implemented
variable 1-4 scaling, which hasn't yet been done. I encourage you to
experiment and check the results carefully. Look at
http://ambermd.org/formats.html for a description of the 1-4 scaling
factors in the topology file, see how it is implemented in Amber (if you
have it available), and then check the values you get from nab vs. the
values you get from sander or pmemd for a variable 1-4-scaled topology file.
One more question, how to revise rgbmax in mmpbsa.py?
>
You will need to write out the mdin files (-make-mdins), change the rgbmax
value in the _MMPBSA_gb.mdin file to what you want, and re-run using
(-use-mdins). Note, if you are using the single trajectory approach for
MMPBSA.py, it does not matter if the 1-4 scaling is handled 'correctly' for
the binding free energy. This is because they will cancel out exactly in
the DELTA term since all conformations will be identical.
HTH,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Sep 27 2012 - 07:30:03 PDT
