Hi Giulia,
My bad. The shell hack I sent you lacked a line that creates the correct
input for sander/pbsa (not tested in mmpbsa_py_energy that you are
using) including:
npbopt=1, eneopt=1, bcopt=5, cutnb=99.0, cutfd=16.0,
cutnb=99.0 is the maximum value it can be in this context. For cutfd I
made several tests and, at least in my case, the energy values converged
at about cutfd=16.0, higher values make the calculations significantly
slower and have very little effect on the results. You may want to see
if that value is a good one in your system.
This is, hopefully, a correct version of the script:
> # Create first the inputs
> MMPBSA.py.MPI -O -i ./binding-${inprot}-${complex}.in -o ${complex}_${post}_${protocol}_mmpbsa.dat -sp ../../${tot}_${post}_sol.prmtop -cp ../../${tot}_${post}_vac.prmtop -rp ../../${rec}_${post}_vac.prmtop -lp ../../${lig}_${post}_vac.prmtop -y ../../${tot}_${post}_sol_prod${prod}_0*mdcrd -make-mdins
> # Non-linear PBSA
> # Add some verbosity
> perl -pi'orig01_*' -e 's|&pb\n|&pb\n npbverb=1, accept=0.000001,\n|' _MMPBSA_${mdinprot}.mdin
> # Make sure inp=1 related options are taken into account
> # Change the PB solver to nonlinear because we have DNA
> perl -pi'orig02_*' -e 's|&pb\n|&pb\n use_sav=0, sprob=1.4, dprob=1.4,\n npbopt=1, eneopt=1, bcopt=5, cutnb=99.0, cutfd=16.0,\n|' _MMPBSA_${mdinprot}.mdin
> # RUN
> MMPBSA.py.MPI -i ./binding-${inprot}-${complex}.in -o ${complex}_${post}_${protocol}_mmpbsa.dat -eo ${complex}_${post}_${protocol}_mmpbsa -sp ../../${tot}_${post}_sol.prmtop -cp ../../${tot}_${post}_vac.prmtop -rp ../../${rec}_${post}_vac.prmtop -lp ../../${lig}_${post}_vac.prmtop -y ../../${tot}_${post}_sol_prod${prod}_0*mdcrd -use-mdins
Cheers,
Miguel
Le 22/03/13 10:18, Giulia a écrit :
> Dear Dr. Jason Swails and Miguel Ortiz Lombardía,
>
> thank you very much for your useful help.
> Regarding the Mg2+ ions:
> as suggested, I will search in the literature if there is any particular prescription.
>
> Regarding the nonlinear PB calculation:
> I run the calculations following the last Miguel's suggestions.
> However, it seems that an other error is present in the calculations.
> Indeed, I had the following error:
>
> Beginning PB calculations with /softmp/AMBER12/amber12/bin/mmpbsa_py_energy
> calculating complex contribution...
> CalcError: /softmp/AMBER12/amber12/bin/mmpbsa_py_energy failed with prmtop TOPO2_DNA_F14.prmtop!
> PB Bomb in pb_read(): cutnb=0 cannot be used with eneopt=1 0.0000000000000000 1
>
>>From the amber12-manual, I read that
> when eneopt=1 (as in my case), cutnb is the cutoff distance used for van der Waals interactions only.
> Moreover, eneopt=1 requires a non-zero CUTNB and BCOPT = 5.
> Now, I am running a test calulation with cutnb=1 (to see if I reach any reasonable result)
> However, in the manual,there is not specified value for cutnb when eneopt=1.
>
> In your experience, do you have any suggestion about the value of cutnb when eneopt=1 ??
>
> Thank you very much
>
> giulia
>
> ________________________________________
> Da: Jason Swails [jason.swails.gmail.com]
> Inviato: giovedì 21 marzo 2013 23.37
> A: AMBER Mailing List
> Oggetto: Re: [AMBER] mmpbsa calculations for a ternary protein/dna/ligand complex in presence of cocrystalized Mg ions
>
> On Thu, Mar 21, 2013 at 11:52 AM, Miguel Ortiz Lombardía <
> miguel.ortiz-lombardia.afmb.univ-mrs.fr> wrote:
>
>>
>> Hi Jason,
>>
>> Well, I did not answer to the Mg2+ question ;-)
>>
>> As for NAB not dealing with non-linear PB, I may be wrong, it may be
>> incompatible with some other option... I carried out many tests and
>> somehow this is what I remember. But I cannot track down now why I think
>> NAB does not deal with non-linear PB, sorry.
>>
>
> Ah, I guess I read too quickly. Well then in response to the Mg2+ question
> from the OP, there is no prescription for dealing with explicit ions in PB
> or GB solvents. You can try running the calculations including whatever
> ions you want to include, but check the results carefully to make sure they
> make sense.
>
> I would also encourage searching the literature for researchers that tried
> something similar to see how they approached the issue.
>
> Good luck,
> 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
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Miguel Ortiz Lombardía
Architecture et Fonction des Macromolécules Biologiques (UMR7257)
CNRS, Aix-Marseille Université
Case 932, 163 Avenue de Luminy, 13288 Marseille cedex 9, France
Tel: +33(0) 491 82 55 93
Fax: +33(0) 491 26 67 20
mailto:miguel.ortiz-lombardia.afmb.univ-mrs.fr
http://www.afmb.univ-mrs.fr/Miguel-Ortiz-Lombardia
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Mar 22 2013 - 04:00:03 PDT
