On Fri, May 14, 2021, Dan Roe wrote:
>
>Unfortunately the 'molsurf' command doesn't have a real per-atom
>surface area calculation - it's one of the long-standing items on my
>to-do list (that only gets longer and longer unfortunately). The LCPO
>surface area calculation ('surf') can do per-atom, but it's quite
>inferior to 'molsurf'.
>
>Maybe there is an enterprising grad. student or post-doc who would
>like to volunteer to add this to CPPTRAJ? :-)
I think some care is required here. The molecular surface, (aka the
solvent-excluded surface, or SES) as defined by Connolly and Richards, and
implemented in molsurf and many other codes, includes surface segments
that can be assigned to a single atom, but also surface segments arising
from places where the probe sphere is in contact with two or more atoms at
the same time. So it's not just a lack of coding that prevents a per-atom
decomposition: one would have to arbitrarily decide how to assign the latter
areas to atoms. And almost any algorithm one thinks of will lead to some
odd behaviors.
One "can" assign surface segments of the solvent-accessible surface (SAS) to
individual atoms. Adding up the values for all atoms in a residue would
then (by one definition) give residue-level values. I'm not sure how useful
such numbers would be, nor whether they would be be less useful than
the corresponding LCPO values.
Long-time participants in this list will know that I am often unhappy with
decomposition schemes for free energies. So don't lose heart too easily:
it's good that there are people working to make good use of such approaches.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri May 14 2021 - 18:30:02 PDT
