[gmx-users] Unnatural amino acid MD

[Mark Abraham]

Ashutosh Jogalekar wrote:
> Hello
> I am trying to do MD on a peptide with an unnatural amino acid, and i  
> have duly modified the OPLS forcefield .hdb, .rtp, and aminoacid.dat  
> file accordingly. Now, when I am trying to do the grompp for the  energy 
> minimization, the program is asking me to put parameters for  the 
> dihedrals in the .top file. I did put most of them in, but cannot  find 
> values for the other ones. More importantly, some of the  dihedrals 
> don't make sense, because they don't seem to have anything  to do with 
> keeping an aromatic ring flat, for example. A few improper  dihedrals 
> should suffice for my amino acid, which is phenylglycine.  But the ones 
> that the program is asking me to put in seem to  correspond to quite 
> obscure atom combinations. Is this a problem with  gromacs, or with the 
> OPLS force field that I am using?

We can't tell without more information. I suggest using the 
phenylalanine entry in the .rtp file as a template to develop your .rtp 
entry. I think the symptoms you are observing are caused by systemic 
problems in what you've tried to do by hand.

Mark