[gmx-users] Diagnosing + system blowing up

[Mark Abraham]

On 30/07/2012 6:36 PM, Shima Arasteh wrote:
> Thanks dear Mark.
> Actually the atomtype C is more suitable for C atom of formyl and not CT1!
>
> The same as before I get the no default bonds/angles/dihedral errors . You explained what I am suppose.
> I can understand the way of matching the coordinate file and atom types in .rtp file. But it's not clear for me that what I am supposed to check? bonds? I don't know what I need to do.

Your .rtp entry has to specify atom types. Those atom types have to make 
chemical sense. Then there have to be bond/angle/dihedral parameters for 
all the combinations you want to use. It may be that this is impossible 
(and pdb2gmx is having trouble, which suggests this is true), but you 
can't know without looking at the force field files. This is about the 
third time I've said to go looking for formyl atom types in 
atomtypes.atp. They're probably not there. Atom types for HC(O)N(H)R... 
are almost certainly not there, and they are what you really want, and 
they'd be quite different from HC(O)R formyl anyway.

> If I look for the bonds/dihedrals and then don't find them, what's the next step?

Search the literature to see if anyone else has used this functional 
group, but probably give up and use acetyl, like everybody else does. 
You have to have a compelling reason to want to bother to parameterize 
this group, and I'd suggest you get a lot more experience first.

Mark