On Mon, 2014-04-14 at 14:50 +0200, Jean-Patrick Francoia wrote:
> Le 14/04/2014 14:10, Jason Swails a écrit :
> > On Sun, 2014-04-13 at 21:40 +0200, Jean-Patrick Francoia wrote:
> >> Hello,
> >>
> >> I am a master student, and one of my course is molecular dynamic. So, in
> >> class we use (the old) Amber 9 to minimize some structures, for example.
> >> Also, one of our exercices is to build the structure of a lysine polymer
> >> with Python (a programmation language), and then to minimize the
> >> structure with Amber.
> >>
> >> The important thing is my Python program generates a SMILE string. I can
> >> import this string with plenty of softwares, and then save it into the
> >> pdb format. But in this pdb, each atom does not have a unique name, a
> >> very common issue when using Amber.
> >>
> >> So, I would like to know if there is a solution to assign each atom a
> >> unique name. I have seen a lot of threads dealing about this problem on
> >> the mailing list, but I have never seen any solution. I can't obviously
> >> rename each atom by hand.
> > LEaP must know about every residue defined in an input structure so it
> > knows what atom types to apply to each atom as well as what charges to
> > apply. To that end, the atom names of every atom must match the names
> > in the library files so that LEaP can map the atom names to the atom
> > types, charges, and connectivities.
> >
> > The atom names of the amino acid and nucleic acid residues match the
> > atom names used by the PDB so that almost any PDB you download from the
> > official Protein Database will work. If you are creating your own PDB
> > from SMILES strings, you will need to fix the atom names yourself. If
> > you are already using Python to do some file manipulations, you can use
> > Python to write a script to rename your atoms.
> >
> > Alternatively, you can use LEaP to create your Lysine polymer with the
> > "sequence" command. If you want a 3-mer of lysine residues, the command
> >
> > lys3 = sequence {NLYS LYS CLYS}
> >
> > will generate a tripeptide with 3 lysine residues.
> >
> > HTH,
> > Jason
> >
>
> I can't use leap to generate the polymer, sadly. Mine is a complex one,
> branched on the epsilon and alpha amines, with several hundreds of amino
> acids.
>
> Ok I get it. But how can I know how the atoms are named in the library ?
> What is the naming algorythm ?
The naming scheme is the one defined by the PDB. You can see what the
names are in the OFF library files inside $AMBERHOME/dat/leap/lib.
>
> And what about generating my own library for my molecule ? Like they do
> here:
> http://ambermd.org/tutorials/basic/tutorial4b/
This is done for individual residues (so only a single amino acid or a
small organic molecule). You will need to break your system down into
smaller parts to parametrize each one. But if you're using amino acids,
why redo all the work that's already been done? Especially since you
need to use the same charge sets that were used to generate the torsion
parameters in the first place.
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Apr 14 2014 - 06:30:02 PDT
