Bug #3213

Having a local residuetypes.dat for custom force field

Added by zhiyi wu over 1 year ago.

Target version:
Affected version - extra info:
Affected version:


I'm trying to port the amber lipid17 force field to Gromcas. However, since the amber lipid17 force field is a modular force field, each lipid consists of three residues, the first acryl chain, the head group and the second acryl chain.
The acryl chain has the same name regardless of whether it is the first acryl chain or the second acryl chain. However, depending on whether the acryl chain is the first or the second, it will have different bonded interactions.
Say we have a DMPC, which will have three residues, MY, PC, MY.
Say if we have atom A in acryl chain 1 (MY) and acryl chain 2 (MY) and atom B in head group (PC). If we want gromacs pdb2gmx to recognise that the first residue which has the residue name of MY is the start of the lipid and the second MY though have the same name is actually the end of the lipid.
In the rtp file, we need to have two different types MY. denoted MY1 and MY2.
where MY1 is

[ MY1 ]
[ atoms ]
A ......
[ bonds ]
A +B

and MY2

[ MY2 ]
[ atoms ]
A ......
[ bonds ]
A -B

In the r2b file, we write

MY    MY   MY1  MY2  -

However, for this to work we need to add the following lines to the residuetypes.dat.

MY1 Protein
MY2 Protein

However, since residuetypes.dat is in the $GROMACSHOME/share/gromacs/top, which is not necessarily accessible to many users using a cluster, I wonder if there is any better way of handling this situation and can one specify the information specified in residuetypes.dat somewhere inside the force field folder?

Also available in: Atom PDF