On Mon, Sep 30, 2013 at 10:27 AM, Aronica, Pietro <
pietro.aronica07.imperial.ac.uk> wrote:
> Hello,
> I need to calculate the interaction between a ligand and a receptor in a
> complex, and I am using MMPBSA to do this.
> My input file is as follows:
>
> &general
> keep_files=1
> /
> &gb
> igb=5, saltcon=0,
> /
> &pb
> istrng=0, inp=1
> /
>
> This will get the GB part running fine, but the PB part will fail citing
> this error:
>
> CalcError: /home/pa07/Amber/amber/bin/mmpbsa_py_energy failed with prmtop
> 144-148_complex.prmtop!
> PB Bomb in pb_aaradi(): No radius assigned for atom 7 CB
> 2C
>
> I have found a workaround on the mailing list, which says to add radiopt=0
> to the PB portion of the input file, which does get the calculation to run.
> However, the result from this calculation is very wrong.
> To compare, here are the GB and PB results:
>
> GB:
> Differences (Complex - Receptor - Ligand):
> Energy Component Average Std. Dev. Std. Err. of
> Mean
>
> -------------------------------------------------------------------------------
> VDWAALS -74.0270 7.7230
> 1.9941
> EEL -110.6010 19.9802
> 5.1589
> EGB 141.0885 14.4495
> 3.7308
> ESURF -10.7044 0.6829
> 0.1763
>
> DELTA G gas -184.6280 17.7664
> 4.5873
> DELTA G solv 130.3841 14.3923
> 3.7161
>
> DELTA TOTAL -54.2439 7.1754
> 1.8527
>
> PB:
> Differences (Complex - Receptor - Ligand):
> Energy Component Average Std. Dev. Std. Err. of
> Mean
>
> -------------------------------------------------------------------------------
> VDWAALS -74.0270 7.7230
> 1.9941
> EEL -110.6010 19.9802
> 5.1589
> EPB 122.2952 15.5015
> 4.0025
> ENPOLAR 1660.6672 80.8117
> 20.8655
> EDISPER 0.0000 0.0000
> 0.0000
>
> DELTA G gas -184.6280 17.7664
> 4.5873
> DELTA G solv 1782.9625 84.6436
> 21.8549
>
> DELTA TOTAL 1598.3345 80.7771
> 20.8566
>
> As you can see, the PB result is not the usual order of magnitude that one
> would expect, whereas the GB one is more in line with what is normally seen.
> This was all done with the latest, completely patched version of Amber.
> Repeating the same calculations with an older version (amber11) gives more
> or less exactly the same GB results (save for a few decimal figures changed
> here and there), while the PB gives a more usual number. This result is the
> same whether radiopt=0 or radiopt=1 is used.
>
> PB (old version calculation)
> Differences (Complex - Receptor - Ligand):
> Energy Component Average Std. Dev. Std. Err. of
> Mean
>
> -------------------------------------------------------------------------------
> VDWAALS -74.0257 7.7225
> 1.9939
> EEL -110.5992 19.9803
> 5.1589
> EPB 122.9205 15.5350
> 4.0111
> ECAVITY -7.9187 0.4359
> 0.1125
>
> DELTA G gas -184.6249 17.7646
> 4.5868
> DELTA G solv 115.0018 15.4784
> 3.9965
>
>
> DELTA G binding = -69.6232 +/- 5.7039
> 1.4727
>
> It seems rather clear to me that there is something wrong with the newer
> version of MMPBSA. Am I using the prmtop incorrectly? Is there a bugfix I
> haven't applied?
> Kind regards
>
The bug you are getting suggests to me that some bugfix is not applied.
(Don't forget that you have to recompile AmberTools after updating!) The
PB energies of the `new' version look good, except for the non-polar
solvation result. Amber 11 uses the standard SASA-based method that GB
uses, which is why the numbers appear 'reasonable'. The latest bug fixes
should have stabilized the issues people were seeing with missing radii for
PBSA.
Another thing you can do is to set molsurf=1 in the &gb section (the latest
version of MMPBSA.py, only) and use the non-polar solvation free energy
computed there in the PB section (this would replace the sum of ECAVITY and
EDISPER).
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Sep 30 2013 - 08:00:04 PDT
