Re: Topology for lysine bouding in Epsilon

From: Kenno Vanommeslaeghe (
Date: Tue May 06 2014 - 13:06:38 CDT

On 05/06/2014 03:35 AM, Jean-Patrick Francoia wrote:
> "That said, your case sounds considerably simpler, so it will probably be
> easier for you to generate each branch as a linear chain of normal
> CHARMM36 RESI LYS monomers, then graft the branches using a 2-residue PRES
> (which you need to put together yourself)."
> I don't really understand the second part of your advice, but what you're
> telling me is basically to minimize each branch normally, and then
> assemble then, am I right ?

No, the [psf] generation step precedes the assignment of Cartesian
coordinates, which in turn precedes any energy calculation, minimization
or simulation. During the generation step, you're working on the abstract
graph that is the connectivity of the molecule, and it is during this step
you can generate [the connectivity for] the linear segments and graft them
together. After that's done, you have to worry about good "initial guess"
Cartesian coordinates (for which I don't have much advice), and *only
then* you can solvate, minimize and simulate.

I always generate my psf using the CHARMM program, so I can give you
detailed instruction on how to use that, but unfortunately, I'm not
confident I can give you good advice for doing the same with psfgen or VMD.

> Another thing I would like to ask is that if I can do what they suggest in
> the doc, for the glutathione (see joinded pdf, page 21). It seems a bit
> hard for me but I can probably manage to do it. If I can't do it, I also
> would like to know why (at least I would have learned something).

This is one of the many possibilities I thought of before writing my last
e-mail; it wouldn't work because glutathione is still linear, whereas you
have a polymer hanging from both your main chain and your side chain.
That's why I recommended you create a 2-residue patch instead. This isn't
covered by that tutorial but it should give you the necessary background
to understand the existing 2-residue patches in the CHARMM36 distribution
and create your own by analogy. I'm not saying it will be easy to do for a
novice, though.

> I will not perform any energy calculations on it,
> I will just use it with a small organic molecule to determine where it
> bounds.

I fail to see how one can determine "where a molecule binds" without
involving energetics of some sort.

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:23 CST