AMBER Archive (2004)

Subject: Re: AMBER: mm_gbsa results for charged ligands

From: Holger Gohlke (gohlke_at_bioinformatik.uni-frankfurt.de)
Date: Fri Dec 24 2004 - 04:31:22 CST

> Here is what I have done to carry out mm_pbsa and
> nmode computations:
>
> First of all, I created AM1-BCC charges for ligand a
> and b in mol2 format by
> antechamber as follows:
>
> Ligand a:
> antechamber -i nea65.amb.pdb -fi pdb -o nea65.mol2 /
> -fo mol2 -at gaff -c bcc -rn NEA -nc 3 -j 5
>
> Ligand b:
> antechamber -i nea74.amb.pdb -fi pdb -o nea74.mol2 /
> -fo mol2 -at gaff -c bcc -rn NEA -nc 2 -j 5

For this limited set of ligands (which are also sufficiently small), I
would have generated charges following the RESP procedure. For one,
AM1-BCC charges are an approximation to RESP. Also - AM1-BCC charges
might not come out right if intra-molecular H-bonds are present in the
structures.

> I then generated prmtop and incrd files for ligand a
> and b in tleap as follows:
>
> tleap -f leaprc.gaff
> nea = loadmol2 nea65(/74).mol2
> set default gibbs off
> set default pbradii mbondi
> saveamberparm nea nea65(/74).prmtop nea65(/74)-start.x
> quit
>
> Afterwards, I pre-minimized the ligands before MD runs
> as follows:
>
> sander -O -i minrelax.in -o nea65(/74).minrelax.out /
> -c nea65(/74)-start.x -p nea65(/74).prmtop /
> -r nea65(/74).minrelax.x
>
> where minrelax.in is
>
> &cntrl
> imin=1, maxcyc=2000, ncyc=250, cut=15, ntpr=50, ntb=0,
> igb=1
> /
>
> After I relaxed the ligands, I executed sander as
> follows
>
> sander -O -i mdnmr.in -o nea65(/74).mdnmr.out /
> -c nea65(/74).minrelax.x -p nea65(/74).prmtop /
> -r nea65(/74).mdnmr.x -x nea65(/74).mdnmr.traj
>
> where mdnmr.in is
>
> &cntrl
> imin=0, ntb=0, igb=1, nstlim=15000, pencut=-0.001,
> nmropt=1, ntpr=10, ntwx=10, t=298.15, ntt=3,
> gamma_ln=1.0,
> tempi=0, temp0=298.15, dt=0.001, cut=15.0, vlimit=5
> /
> &wt type='REST', istep1=0,istep2=5000,value1=0.1,
> value2=1.0, /
> &wt type='REST', istep1=5001,istep2=15000,value1=1.0,
> value2=1.0, /
> &wt type='END' /
>
> LISTOUT=POUT
> DISANG=RSTpH65(/74)
>

The above seems ok.

> I then did GB stability computations for the final
> coordinates of
> the MD runs [nea65(/74).mdnmr.x].

First, I would average stability computations over an ensemble of
structures instead of using just the final snapshots. Effective energies
can show quite some fluctuations depending on the molecular structure
used, so averaging should provide an answer "closer to the expected
value" (and it also gives you an idea about your errors in the
calculations). Second, GB calculations are an approximation to PB
calculations. For your system size, one can easily afford PB
calculations.

> Here are the nmode results for ligand a and b;
>
> Ligand a:
>
> E Cv S
> kcal/mol cal/mol-kelvin cal/mol-kelvin
> ------------------------------------------------------
> Total 291.251 94.421 226.691
> translational 0.888 2.979 43.206
> rotational 0.888 2.979 34.835
> vibrational 289.475 88.463 148.651
>
> Ligand b:
>
> E Cv S
> kcal/mol cal/mol-kelvin cal/mol-kelvin
> ------------------------------------------------------
> Total 283.197 92.348 183.544
> translational 0.888 2.979 43.197
> rotational 0.888 2.979 34.765
> vibrational 281.421 86.390 105.582
>
> As seen above entropic contributions (Total S) to
> absolute G are
> favorable for both Ligand a and b. That is fine!
> Then, why do I get an unfavorable GBTOT value for
> Ligand b ?

I don't see your point here. "Absolute" entropies are always positive
and, hence, they will always contribute favorably to G, so to me this
does not provide a criterion as to whether your structures are
"meaningful/favorable".

Best regards

Holger 

-- 
++++++++++++++++++++++++++++++++++++++++++++++++++
Dr. Holger Gohlke

J.W. Goethe-Universität
Fachbereich Biologie und Informatik
Institut für Mikrobiologie	
Marie-Curie-Str. 9              
60439 Frankfurt/Main            
Germany                         

Tel.:  (+49) 69-798-29411; Fax: (+49) 69-798-29826
Email: gohlke_at_bioinformatik.uni-frankfurt.de
URL:   http://www.uni-frankfurt.de/~hgohlke
++++++++++++++++++++++++++++++++++++++++++++++++++
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu