On Sat, Jan 07, 2017, David Poole wrote:
>
> Now, having established the problem, I'd like to ask another question
> in general about tleap, specifically, how does it assign the location
> of missing atoms for library or added molecules?
It generates internal coordinates from those in the library file; then uses
those to create internal coordinates for the missing atoms in the PDB file;
then converts everything back to cartesians.
This is pretty unsophisticated, but the hope is that miminization and MD will
"fix" any problems arising from the initial structures. Reduce will generally
do a better job of adding missing hydrogens. If you have missing side chains
(or parts of sidechains) there are a variety of programs that will use
rotamer libraries and other information to create better starting structures.
Bottom line (for me) if you have only a few missing atoms, the what tleap
does is probably fine. But there are often better ways to get good starting
structures--you need to consider how important a good starting structure is to
your final result. (For example, if you plan to do lots of sampling, say via
replica exchange, then details of the starting structure may not be very
important. If you are planning just to do conventional MD, then trying hard
to avoid bad starting configurations is probably worth the trouble, since
getting "stuck" in a bad starting configuration is then much more likely.)
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Jan 07 2017 - 15:00:03 PST
