[gmx-users] "Fatal error: Atomtype CH2 not found" in grompp command

[Justin Lemkul]

On 8/16/16 3:04 PM, Elissa Fink wrote:
> Dear Gromacs users,
>
> I am simulating a protein-ligand complex and I am performing the grompp
> command before adding ions to the system:
>> gmx grompp -f em.mdp -c solv.gro -p topol.top -o ions.tpr
>
> and I get the error:
>
> "Source code file:/build/gromacs-z6bPBg/gromacs-5.1.2/src/gromacs/gmxpreprocess/toppush.c,
> line:1353
>
> Fatal error:
> Atomtype CH2 not found"
>
> I checked for CH2 groups in the solv.gro file and there are 4 in tryptophan

Anything that says CH2 in a coordinate file is an atom name, not an atom type. 
The problem comes from atom types, which is a force field/topology issue.

> residues. I am running the complex with the amber99sb-ildn force field. I
> checked the force field .rtp file for the tryptophan name of a CH2 group.
> It has CA, so I changed the 4 CH2 groups to CA in the solve.gro file. I
> reran the grompp command above and got the same error. This makes me think
> the problem is not the tryptophan CH2's in the solve.gro file.
>
> Any recommendations for addressing the CH2 atom type not found?
>

A CH2 type is a united-atom type coming from the GROMOS force fields.  This 
means you've mixed and matched force fields, which is not appropriate.  You need 
to have a self-consistent force field representation for the system.

-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

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