[gmx-users] Regarding non zwitterionic structure to zwitterionic structure while using gmxpdb2gmx

[Justin Lemkul]

On 10/18/17 6:05 AM, Dilip H N wrote:
> Hello,
> I have a neutral glycine molecule [NH2CH2COOH] and if i give gmx pdb2gmx i
> m getting error as:-
> Fatal error:
> Atom HO in residue GLY 1 was not found in rtp entry GLY with 11 atoms
> while sorting atoms.
> For a hydrogen, this can be a different protonation state......
>
> So i use the -ignh flag (gmx pdb2gmx gly.pdb -o gly.gro -ignh) which gives
> me a zwitterionic structure and topology is created.(since i need to
> simulate in water, the zwitterionic structure is good.)
>
> 1] But if i convert the zwitterionic form of glycine into .mol2 format and
> upload to CGENFF server (i am using charmm FF) for creating the topology,
> the .str file that is generated shows a penalty of 99.5, which is not good.
>
> RESI glycin       0.000 ! param penalty=   0.100 ; charge penalty=  99.591
> GROUP            ! CHARGE   CH_PENALTY
> ATOM N      NG3P3  -0.336 !   32.195
> ATOM H1     HGP2    0.330 !    2.200
> ATOM H2     HGP2    0.330 !    2.200
> ATOM H3     HGP2    0.330 !    2.200
> ATOM CA     CG324   0.241 !   99.591
> ATOM HA1    HGA2    0.090 !    3.821
> ATOM HA2    HGA2    0.090 !    3.821
> ATOM C      CG2O3   0.445 !   98.841
> ATOM OT1    OG2D2  -0.760 !    8.256
> ATOM OT2    OG2D2  -0.760 !    8.256
>
> But if i compare the topology created by 'gmx pdb2gmx gly.pdb -o
> gly.gro -ignh' command then the charges of the atoms are different...
> then how is this possible..??

CGenFF parametrizes whatever it finds based on a set of rules, which 
come from relationships to other molecules. This is intrinsically 
different than the library of highly optimized residues already built 
into the force field.

DO NOT use CGenFF topologies for molecules that are already specifically 
tuned as part of the CHARMM force field. Use what pdb2gmx gives you.

-Justin

> Which topology is more reliable..??
>
> 2] if non zwitterionic form of glycine is given in form of .mol2, then
> the .str file give zero penalty.
>
> RESI *****        0.000 ! param penalty=   0.000 ; charge penalty=   0.000
> GROUP            ! CHARGE   CH_PENALTY
> ATOM N      NG321  -0.956 !    0.000
> ATOM CA     CG321   0.100 !    0.000
> ATOM C      CG2O2   0.751 !    0.000
> ATOM O      OG2D1  -0.551 !    0.000
> ATOM HA1    HGA2    0.090 !    0.000
> ATOM HA2    HGA2    0.090 !    0.000
> ATOM H      HGPAM2  0.338 !    0.000
> ATOM HN     HGPAM2  0.338 !    0.000
> ATOM OXT    OG311  -0.629 !    0.000
> ATOM HO     HGP1    0.429 !    0.000
>
> So which topology is more reliable for simulations..??
>
> Any suggestions ar appreciated..
>

-- 
==================================================

Justin A. Lemkul, Ph.D.
Assistant Professor
Virginia Tech Department of Biochemistry

303 Engel Hall
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.biochem.vt.edu/people/faculty/JustinLemkul.html

==================================================