[gmx-users] No default Proper Dih. types when using neutral terminal

[Justin Lemkul]

On 10/26/17 10:09 AM, Simon Kit Sang Chu wrote:
> Hi everyone,
>
> Last time, I mentioned missing parameter for neutral terminal dihedral and
> U-B type. After some debugging, I am still not sure what the cause was. I
> hope providing more information could help to resolve the issue. I am using
> CHARMM36 forcefield.
>
> Brief summary of the problem -
> No dihed. parameter / U-B types after using neutral terminals
>
> Command I ran -
>
>     1. gmx pdb2gmx -f ALA-GLY-GLY-PRO. pdb -o ALA-GLY-GLY-PRO.gro -p
>     ALA-GLY-GLY-PRO.top -i ALA-GLY-GLY-PRO.itp -water tip3p -ignh -ter
>     Choose CHARMM36 all-atom force field (November 2016), NH2 and COOH
>     terminals

Note that the combination of NH2 and COOH cannot occur at any real pH 
value, so choose wisely what you're doing...

>     2. gmx editconf -f ALA-GLY-GLY-PRO.gro -o ALA-GLY-GLY-PRO-box.gro -c -d
>     0.8 -bt dodecahedron
>     3. gmx solvate -cp ALA-GLY-GLY-PRO-box.gro -cs spc216.gro -o
>     ALA-GLY-GLY-PRO-solvated.gro -p ALA-GLY-GLY-PRO.top
>     4. gmx grompp -f ions.mdp -c ALA-GLY-GLY-PRO-solvated.gro -p
>     ALA-GLY-GLY-PRO.top -o ions.tpr
>
> Fatal error in grompp -
>
>
> ERROR 1 [file ALA-GLY-GLY-PRO.top, line 293]:
>    No default U-B types
>
> ERROR 2 [file ALA-GLY-GLY-PRO.top, line 367]:
>    No default Proper Dih. types
> Meanwhile in the topology file -
>
> [ angles ]
>
>     30    38    40     5          ( line 293 )
>
>
> [ dihedral ]
>
>     27    26    30    38     9  ( line 367 )

These are indeed missing parameters in the CHARMM force field, derived 
from the fact that the CA atom type is different in PRO than any other 
amino acid. Probably no one has ever needed to do a simulation of a 
system like this so no one ever ran into it. You can probably safely 
assign parameters by analogy using anything that has CT1 instead of CP1, 
which is the CA type in PRO.

-Justin

> ; residue   5 PRO rtp PRO  q  0.0
>      26          N      5    PRO      N     26      -0.29     14.007   ;
> qtot -0.29
>      27        CP3      5    PRO     CD     27          0     12.011   ;
> qtot -0.29
>      28        HA2      5    PRO    HD1     28       0.09      1.008   ;
> qtot -0.2
>      29        HA2      5    PRO    HD2     29       0.09      1.008   ;
> qtot -0.11
>      30        CP1      5    PRO     CA     30       0.02     12.011   ;
> qtot -0.09
>      31        HB1      5    PRO     HA     31       0.09      1.008   ;
> qtot 0
>      32        CP2      5    PRO     CB     32      -0.18     12.011   ;
> qtot -0.18
>      33        HA2      5    PRO    HB1     33       0.09      1.008   ;
> qtot -0.09
>      34        HA2      5    PRO    HB2     34       0.09      1.008   ;
> qtot 0
>      35        CP2      5    PRO     CG     35      -0.18     12.011   ;
> qtot -0.18
>      36        HA2      5    PRO    HG1     36       0.09      1.008   ;
> qtot -0.09
>      37        HA2      5    PRO    HG2     37       0.09      1.008   ;
> qtot 0
>      38         CD      5    PRO      C     38       0.72     12.011   ;
> qtot 0.72
>      39         OB      5    PRO    OT1     39      -0.55    15.9994   ;
> qtot 0.17
>      40        OH1      5    PRO    OT2     40      -0.61    15.9994   ;
> qtot -0.44
>      41          H      5    PRO    HT2     41       0.44      1.008   ;
> qtot 0
>
>
> I further check if angle type between CP1, CD and OH1 exists in ffbond.itp.
> Indeed there is no such angle.  Maybe there is something wrong when using
> NH2 and COOH terminals?
>
> Please feel free to give any suggestion. I appreciate any comment.
>
> Regards,
> Simon

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

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

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