On Sat, Nov 12, 2005, Yong Duan wrote:
>
> Can somebody out there tell me how can I
> add a pair of atoms on the excluded atom list and expect sander to use
it?
A little trial-and-error is probably appropriate here. As suggested
earlier
in this thread, it looks like setting chngmask to 0 should turn off any
new
excluded-list generation.
Looking the at the code, I see we already trap the combination of LES and
extra points, printing out an error message that this is not supported.
If
LES is set and there are no extra points, then the exclusion list is *not*
remade -- which is presumably why all the LES tests still pass.
Overall, Tom did a pretty reasonable job of being careful in inserting the
new
code: he trapped the known options that require one to use the excluded
list
from the prmtop file (rather than rebuilding it), and he added an input
option
that disables the rebuilding anyway. A few more comments might have
helped,
but I'm hardly in a position to complain about that...
[A question for Tom D.: is there a reason why we don't skip all the
extra-points initialization whenever numextra = 0? That is have a line:
if( numextra == 0 ) return
at about line 198 of extra_pts.f? I know that this would require another
change in the ephi() routine, but otherwise seems pretty simple. The
objective would be to avoid running code that is not needed.]
...dac
Received on Wed Apr 05 2006 - 23:49:48 PDT
