[gmx-users] pdb2gmx and specbond with arbitrary ligand

[Leandro Bortot]

Hi,

     Thank you Justin. I added the ligand to the forcefield as a new
residue by editing local copies of aminoacids.rtp, atomtypes.atp,
ffbonded.itp and ffnonbonded.itp using information from the .itp for the
isolated ligand I already had. Additionally, I created the protein-ligand
bonds using specbond.dat. Everything worked well.





     For the sake of helping others who may encounter similar situations, I
report two things I didn't expect:


     First: ffbonded.itp is not case sensitive. I had the new atom types
"c" and "o" and grompp gave me the following warnings:

"











*WARNING 1 [file ffbonded.itp, line 116]:  Overriding Bond parameters.
old:                                          0.1229 476976 0.1229 476976
new: c   o        1    0.12140   542250.0WARNING 2 [file ffbonded.itp, line
404]:  Overriding Angle parameters.
old:                                          126 669.44 126 669.44   new:
o    c    o   1    130.380    654.130*
"

     The old values are the parameters for the the "C O" bond and "O C O
" angles, which are conflicting with the new "c o" bond and "o c o" angle.
I solved this by renaming "c" and "o" to something else, but I didn't
expected this to happen since ffnonbonded.itp is case sensitive.



     Second: the AMBER99SB-ILDN forcefield uses the dihedral function 9 for
propers and function 4 for impropers. On the other hand, the
parametrization by GAFF/Antechamber/acpype uses function 3 for propers and
function 1 for impropers. The problem is that pdb2gmx seems to assume that
all proper dihedrals use function 9 and impropers use function 4, including
the ligand ones. As a result, grompp give many "No default Proper Dih.
types" and "No default Improper Dih. types" errors. Since the ligand is in
the end of the chain .itp file that pdb2gmx created, this is easily fixed
by renumbering the adequate dihedral functions in a text editor.




Thank you for your attention,
Leandro.



On Wed, Mar 11, 2015 at 10:20 PM, Justin Lemkul <jalemkul at vt.edu> wrote:

>
>
> On 3/11/15 1:08 PM, Leandro Bortot wrote:
>
>> Dear users,
>>
>>       I want to simulate some ligands covalently bonded to a protein, i.e.
>> covalent inhibitors. I've searched the list and found similar issues, but
>> mainly with heme in forcefields in which it is already supported. I use
>> the
>> AMBER99SB-ILDN forcefield and I can simulate the non covalent complexes
>> with no problems using GAFF.
>>
>>       I suppose I can just put both the ligand and the protein under the
>> same [moleculetype] specification in a .itp file and manually add the
>> bonds
>> in this topology. It seems that I only need a script to renumber the atoms
>> to achieve this. However, I think pdb2gmx and specbond.dat may already be
>> able to do this. The problem is: How can pdb2gmx understand my arbitrary
>> ligand? Is adding my ligand to a local copy of
>> amber99sb-ildn/aminoacids.rtp the only way?
>>
>>       What would be the best way to do that?
>>
>>
> Yes, the .rtp approach is the most reliable way to do this.  Follow all
> the steps in http://www.gromacs.org/Documentation/How-tos/Adding_
> a_Residue_to_a_Force_Field and add an entry in specbond.dat (format
> explained in http://www.gromacs.org/Documentation/File_Formats/
> specbond.dat)
>
> -Justin
>
> --
> ==================================================
>
> Justin A. Lemkul, Ph.D.
> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>
> Department of Pharmaceutical Sciences
> School of Pharmacy
> Health Sciences Facility II, Room 629
> University of Maryland, Baltimore
> 20 Penn St.
> Baltimore, MD 21201
>
> jalemkul at outerbanks.umaryland.edu | (410) 706-7441
> http://mackerell.umaryland.edu/~jalemkul
>
> ==================================================
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/
> Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>