Dear Jason,
I have checked and I don't see any difference in number of atoms. I have added files to dropbox for close inspection
https://www.dropbox.com/sh/zb8flt93fd6mn3g/RMcZ8NkGRp/MMPBSA%20ERROR
regards,
Nitin Sharma
-----Original Message-----
From: Jason Swails [mailto:jason.swails.gmail.com]
Sent: Thursday, April 24, 2014 10:47 PM
To: amber.ambermd.org
Subject: Re: [AMBER] AMBER minimization problem
On Thu, 2014-04-24 at 13:52 +0000, Valentina Romano wrote:
> Hi
>
> I minimized the 6-aminopurine itself twice.
>
> 1) the frcmod (generated by parmchk) was:
>
> remark goes here
> MASS
>
> BOND
>
> ANGLE
>
> DIHE
>
> IMPROPER
> ca-cc-na-hn 1.1 180.0 2.0 General improper torsional angle (2 general atom types)
> h5-na-cc-nd 1.1 180.0 2.0 Using default value
> ca-ca-ca-nd 1.1 180.0 2.0 Using default value
> ca-nb-ca-nh 1.1 180.0 2.0 Using default value
> ca-hn-nh-hn 1.1 180.0 2.0 Using default value
> h5-nb-ca-nb 1.1 180.0 2.0 Using default value
> ca-na-ca-nb 1.1 180.0 2.0 Using default value
>
> NONBON
>
> Then, I minimized the ligand in vacuum and without periodicity and the molecule was not planar any more.
>
> 2) I increased the force constants for improper angles and thus the frcmod file was:
>
> remark goes here
> MASS
>
> BOND
>
> ANGLE
>
> DIHE
>
> IMPROPER
> ca-cc-na-hn 5.6 180.0 2.0 General improper torsional angle (2 general atom types)
> h5-na-cc-nd 5.6 180.0 2.0 Using default value
> ca-ca-ca-nd 5.6 180.0 2.0 Using default value
> ca-nb-ca-nh 5.6 180.0 2.0 Using default value
> ca-hn-nh-hn 5.6 180.0 2.0 Using default value
> h5-nb-ca-nb 5.6 180.0 2.0 Using default value
> ca-na-ca-nb 5.6 180.0 2.0 Using default value
>
> NONBON
>
> Then I ran a minimization as before and the molecule had a planar
> conformation. So it look like a problem was solved
>
> Then i got an additional problem:
> I created topology and coordinate files for the 6-aminopurine-PknG complex (following the steps of the tutorial B4) and of course i used the ligand's frcmod file manually modified (which worked for the ligand alone).
> Then i minimized the complex:
>
> Initial minimisation of PknG-Adenine complex &cntrl
> imin=1,
> maxcyc=500,
> ncyc=250,
> ntb=0,
> igb=0,
> cut=12
> /
>
> After the minimization the 6aminopurine within the binding pocket had
> lost again is planarity. It had a distorted conformation.
>
> Do you have any idea of how i could avoid the ligand distortion when i
> minimized the whole complex and not only the ligand itself?
So you increased the force constant from 1.1 to 5.6 for all of the improper torsions (which are the parameters typically used to maintain planarity). This kept the isolated ligand planar during a minimization.
This means that there were _some_ forces (probably non-bonded forces) that were pushing the molecule away from planarity harder than the impropers were trying to keep the molecule planar. When you made the impropers stiffer, they were able to overpower the effects that were resisting planarity.
When you embed it in the system, it loses its planarity again. This indicates that there are now even _stronger_ forces pushing against planarity than there were outside of the bound complex that can now overcome the strength of the 'stronger' improper terms. It stands to reason that making the force constant even stronger will keep the rings planar more.
Force fields are remarkably simple -- there is a simple functional form for the interaction of particles from which the reasoning I used above naturally follows.
However, adjusting the force field to get the results you expected to get in the first place is bad science in my opinion. Force fields are parametrized against experimental measurements or higher-level quantum mechanical calculation. Unless you can justify your changes to the parameters with something besides "I think this molecule should be planar" there is no reason to trust your new parameter set. (And as Dave mentioned before, assuming that your 6-amino purine is planar in a bound complex is not necessarily accurate -- it could very well be non-planar when bound in a complex).
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
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 Thu Apr 24 2014 - 09:00:02 PDT
