Hi Jason,
> I have a slight concern regarding variable 1-4 scaling as it's
> implemented
> in tleap. Looking at one of the topology files created by tleap, it
> appears
> as though several dihedral 1-4 scaling parameters are just set to 0 in
> the
> topology file. However, it also appears these may correspond to
> ignored
> scaling factors (as with multi-term dihedrals and impropers, etc.), so
> they
> don't really affect results (hopefully). Indeed in a small test case,
> hard-coding all of the 1-4 scaling constants to be 1.2 and 2.0 appears
> to
> leave the results unchanged wrt the old topology file.
>
> They are inverted as soon as they're read in, though, so "0" may not be
> the
> most industrious of choices here... Is there something I'm missing?
This was my suggestion and adding the 0 to ignored dihedrals knowing it
would be inverted was the whole point. That way if anything is messed up in
the code for calculating the 1-4 scaling or something modifies the list in a
correct way you will end up with infinite energies and forces for the 1-4
EEL and VDW terms and the calculation should stop. Rather than producing
'slightly' wrong results.
It also makes it easier to see which dihedrals are duplicates or not.
I guess the only issue is if a compiler objects to a division by zero but my
understanding is that division by a floating zero should not throw and
error, only integer division by zero should do this.
All the best
Ross
/\
\/
|\oss Walker
---------------------------------------------------------
| Assistant Research Professor |
| San Diego Supercomputer Center |
| Adjunct Assistant Professor |
| Dept. of Chemistry and Biochemistry |
| University of California San Diego |
| NVIDIA Fellow |
|
http://www.rosswalker.co.uk |
http://www.wmd-lab.org/ |
| Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
---------------------------------------------------------
Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Fri Aug 12 2011 - 10:00:05 PDT
