I see and thank you. So is there a way to retain/report the internal
energy values for a run like:
nohup mpirun -np 16 $AMBERHOME/bin/MMPBSA.py.MPI -O -i mmpbsa.in -o
FINAL_RESULTS_MMPBSA.dat -sp prmtop -cp com.prmtop -rp rec.prmtop -lp
lig.prmtop -y ../*.mdcrd > benchmarks.log &
such as get a similar report for all E terms like the complex only run
where the lig, rec and comp all show the internal energy contributions
even though they should cancel out upon taking the difference.
the thing is we might need them for comparison for other runs.. please
advise.
Dan
On Thu, April 21, 2016 9:12 pm, Jason Swails wrote:
> On Thu, Apr 21, 2016 at 3:42 PM, McElheny, Dan <dmcelh1.uic.edu> wrote:
>
>> Hi,
>> I'm trying to determine the Energy of the complex only using mmpbsa. in
>> doing so i'm comparing 2 runs where:
>>
>> #mmpbsa.in
>> nput file for running PB and GB
>> &general
>> verbose=1,
>> receptor_mask=':1-192',
>> ligand_mask=':193-384'
>> /
>> &gb
>> igb=5, saltcon=0.100
>> /
>> &pb
>> istrng=0.100,
>> radiopt=0
>> /
>>
>> Looking at the Energies I get two very different results and was
>> wondering
>> if they should be the same actually.
>>
>> 1) the ligand, receptor and complex are all included.
>>
>> nohup mpirun -np 16 $AMBERHOME/bin/MMPBSA.py.MPI -O -i mmpbsa.in -o
>> FINAL_RESULTS_MMPBSA.dat -sp prmtop -cp com.prmtop -rp rec.prmtop -lp
>> lig.prmtop -y ../*.mdcrd > benchmarks.log &
>>
>> GENERALIZED BORN:
>>
>> Complex:
>> Energy Component Average Std. Dev. Std. Err.
>> of
>> Mean
>>
>> -------------------------------------------------------------------------------
>> VDWAALS -2772.0879 64.4476
>> 2.8822
>> EEL -26117.3304 242.6761
>> 10.8528
>> EGB -7799.0717 244.1382
>> 10.9182
>> ESURF 112.7079 6.8321
>> 0.3055
>>
>> G gas -28889.4183 199.5336
>> 8.9234
>> G solv -7686.3638 248.8984
>> 11.1311
>>
>> TOTAL -36575.7821 76.8913
>> 3.4387
>>
>
> ​Internal terms that cancel out in a single-trajectory approach are
> omitted, since they contribute exactly 0 to the binding energy (this is
> actually checked and if they do not cancel, is reported as a likely
> error). So bond, angle, torsion, and 1-4 nonbonded terms are explicitly
> excluded from the total here.
> ​
>
>
>>
>> 2) only the complex is flagged
>>
>> nohup mpirun -np 16 $AMBERHOME/bin/MMPBSA.py.MPI -O -i mmpbsa.in -o
>> FINAL_RESULTS_MMPBSA.dat -sp prmtop -cp com.prmtop -y ../*.mdcrd >
>> benchmarks.log &
>>
>>
>> GENERALIZED BORN:
>>
>> Complex:
>> Energy Component Average Std. Dev. Std. Err.
>> of
>> Mean
>>
>> -------------------------------------------------------------------------------
>> BOND 1056.8691 25.5803
>> 1.1440
>> ANGLE 3229.6268 43.1459
>> 1.9295
>> DIHED 4289.7506 23.2319
>> 1.0390
>> VDWAALS -2698.5830 66.0139
>> 2.9522
>> EEL -26401.6751 173.8585
>> 7.7752
>> 1-4 VDW 1259.7356 14.1568
>> 0.6331
>> 1-4 EEL 21641.9617 42.1481
>> 1.8849
>> EGB -7567.5331 162.8158
>> 7.2813
>> ESURF 121.7999 7.5300
>> 0.3368
>>
>> G gas 2377.6856 172.8013
>> 7.7279
>> G solv -7445.7332 164.8591
>> 7.3727
>>
>> TOTAL -5068.0476 73.0275
>> 3.2659
>>
>
> ​For a single snapshot, internal terms *don't* cancel, so you get the
> bond,
> angle, dihedral, and 1-4 terms included in the total here. So obviously
> the energy will be very different. However, as long as you use exactly
> the
> same prmtop and trajectory files (and input files), and analyze exactly
> the
> same frames, then the terms printed in both output files should match.
> They're close in this example, but are certainly not exactly matching, but
> it's not clear to me if that's expected or not...
>
> HTH,
> Jason
>
> --
> Jason M. Swails
> _______________________________________________
> 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 Fri Apr 29 2016 - 14:00:03 PDT
