[gmx-users] topology file

[Erik Lindahl]

Shrivastava, Indira (NCI) wrote:
>  Hi All, 
> 
>  I would like to make a 'top' file for a lysolipid (lysophosphatidylcholine
> (L-POPC))
>  This lipid is a single chain lipid, with the same headgroup as
> phosphatidylcholine (POPC).
>  Can I use the forcefield parameters of POPC to create a top file for
> (L-POPC) since they
>  are essentially identical, the only difference being in POPC there are two
> chains whereas in 
> L-POPC there is only one chain. 
> 

Yes, that's what I would do. Lipid topologies are quite easy to edit 
manually even if it might seem hard the first couple of times.

If possible, keep the headgroup and the first chain to avoid renumbering 
any atoms, and then edit the remaining bonds/angles/dihedrals if you 
need to.

To test things, run a short simulation of a single lipid in vacuum and 
check it with ngmx. Since ngmx shows the real bonds defined in the tpr 
file (in contrast to e.g. rasmol which draws a bond whenever two atoms 
are close) you can see if the connectivity is correct. You also want to
make sure that the energies for angles and dihedrals are reasonable
for the corresponding number of degrees of freedom.

Cheers,

Erik