[gmx-users] Re: phosphoserine answer (and phosphate related question)

[Graham Smith]

> I am having trouble to generate a topology file for a protein which
> contains a phosphoserine residue;  Have anyboby already worked with
> it? 

I made a phosphoserine a little while ago; you can get it from 
"ffG43ap" in the topologies/forcefields section of gromacs.org.
I don't think it's too bad, though I've only really used the
phosphothreonine. 

I've been working on making ADP/ATP/GDP/GTP nucleotides recently too. 
I'm finding I cant' stop the triphosphate groups getting very
distorted with an MG ion there (like an umbrella turning inside out) -
is this normal with highly charged species? At the moment the ion is a
separate molecule; would it be better to bond it explicitly into the
molecule so I can put in Mg-O-P angle terms etc?  I'd be interested to
hear if anyone else tried this.