All,
Thank you for the input.
This was operator error on my part and has been fixed.
What I should have done was to create the desired ligand parameters from
the atoms and coordinates in the PDB file of the complex + ligands and then
generate the .prepi and .frcmod files in the usual way; what I had done was
to create the .prepi and .frcmod files from atoms and coordinates from a
SDF file for said ligand.
This resulted in a mismatch of atom numbers and connectivity when parsing
the PDB file, although at the end of the day it gave me an unintended but
desired simulation where I have the receptor of interest with one S in one
binding pocket and one R in the other.
How exactly do you get it right?
(1) grep LIG PDB.pdb > LIG_from_PDB.pdb, where the string "LIG" will
certainly be something different but is the ligand of interest
(2) use pymol (or your favorite) to add a hydrogen in the right place,
creating the +1 S enantiomer; export as LIG_S_charged.pdb
(3) split your terminal window and - using only vi, of course - open
PDB.pdb on the left and LIG_S_charged.pdb on the right
(4) make sure the atom names and numbers from the left pane (e.g., C1 and
N2) coordinates match (or very closely) to those in the right panel
(5) having confirmed this, number the atoms on the right (in
LIG_S_charged.pdb) accordingly - if done correctly, you will only have to
add numbers on the hydrogens
(6) antechamber -fi pdb -i LIG_S_charged.pdb -fo prepi
-o LIG_S_charged.prepi -seq yes -nc 1 -c bcc -rn LIG -j 5
(7) parmchk2 -f prepi -i LIG_S_charged.prepi -o LIG_S_charged.frcmod
(8) in tleap:
source leaprc.protein.ff19SB
source leaprc.gaff
source leaprc.water.opc
loadamberprep LIG_S_charged.prepi # Load the prepi file for the ligand
loadamberparams LIG_S_charged.frcmod # Load the frcmod file for ligand
mol = loadpdb PDB.pdb # Load PDB file for protein-ligand complex
solvatebox mol TIP3PBOX 8 # Solvate the complex with a cubic water box
addions mol Cl- 0 # Add Cl- ions to neutralize the system
addions mol Na+ 0 # Add Na+ ions to neutralize the system
saveamberparm mol complex_initial.prmtop complex_initial.crd
(9) for me, I'm using OpenMM for the MD simulation engine and AmberTools
for everything else - run minimization/MD and analyze using any number of
tools including cpptraj/pytraj/mdtraj/etc.
(10) write all this down and make sure you don't mess it up the next time.
This is a bit of a rookie mistake, and I'm chalking this up to my having
been away from AMBER-world for 17 years - but now I'm back.
Thanks again y'all,
Todd Minehardt
On Tue, Sep 5, 2023 at 6:30 PM Carlos Simmerling via AMBER <
amber.ambermd.org> wrote:
> I'm still not quite clear - what comes out of leap before anything else?
> The atoms should match the input pdb, except for missing atoms (and then
> you should get a warning in the leap log).
>
> On Tue, Sep 5, 2023 at 6:43 PM Todd Minehardt <tminehardt.nobiastx.com>
> wrote:
>
> > Carlos,
> >
> > I've got a case where the neutrally-charged ligand heavy atoms are
> present
> > in the PDB and are from the neutral S-enantiomer that was crystalized
> > during experiment.
> >
> > I'm generating the +1 charged S enantiomer as prepi and frcmod files as
> > described, and then assuming (perhaps incorrectly) that for each of the 2
> > occurrences of LIG in the PDB that the requisite number of hydrogens -
> > including the one that adds the +1 and the one that makes it the S
> > enantiomer - get added.
> >
> > The initial structure (after solvation, neutralization, etc.) shows that
> > the geometry of each of the LIGs is not quite right, but shakes out to
> one
> > S and one R after minimization.
> >
> > Keep in mind that the LIG atoms are numbered in the PDB, and I'm going
> > into this after removing CONECTs and inserting a TER card between the 2
> > LIGs.
> >
> > I've already had issues with the GAFF atom types - which is expected -
> but
> > am baffled to why, given one S enantiomer prepi/frcmod combo generated
> from
> > the all-atom SDF from PubChem, the two LIGs shake out, as it were, to
> one S
> > and one R.
> >
> > Note: the same system except with neutrally-charged S LIGs worked just
> > fine - 2 starting S, 2 ending S after 100 ns.
> >
> > Help is much appreciated,
> >
> > Cheers,
> >
> > Todd
> > ------------------------------
> > *From:* Carlos Simmerling <carlos.simmerling.gmail.com>
> > *Sent:* Tuesday, September 5, 2023 4:23 PM
> > *To:* Todd Minehardt <tminehardt.nobiastx.com>; AMBER Mailing List <
> > amber.ambermd.org>
> > *Subject:* Re: [AMBER] Chirality flipped for 2 identical molecules
> >
> > I don't think leap should change this, what enantiomers are in the input
> > pdb? Did leap change them? If so, which atoms got moved?
> >
> > On Tue, Sep 5, 2023, 4:41 PM Todd Minehardt via AMBER <amber.ambermd.org
> >
> > wrote:
> >
> > I am constructing prepi and frcmod files from a PubChem 3D sdf file for a
> > positively-charged (+1) S enantiomer. I do this by running antechamber
> and
> > parmchk2:
> >
> > antechamber -fi sdf -i compound.sdf -fo prepi -o compound.prepi -nc 1
> -seq
> > y -j 5 -rn LIG -c bcc
> >
> > parmchk2 -f prepi -i compound.prepi -o compound.frcmod
> >
> > I have a nice clean PDB structure in which there are two occurences of
> the
> > heavy atoms of LIG.
> >
> > When I run this through tLEaP, one LIG is S and the other is R (after
> > using ambpdb for generating PDB from saved .prmtop and .crd files).
> >
> > Is there an explanation for this behavior, as it is repeatable; and I'm
> > not doing anything at all to the compound.sdf file and only removed
> > unwanted waters and the like from the PDB file from RCSB.
> >
> > Thank you,
> >
> > Todd Minehardt
> > _______________________________________________
> > 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
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Sep 25 2023 - 07:30:02 PDT
