From: Jérôme Hénin (jhenin_at_ifr88.cnrs-mrs.fr)
Date: Mon Dec 12 2011 - 11:30:47 CST
On 12 December 2011 17:13, Jian Dai <jdai2_at_fsu.edu> wrote:
> Dear all:
> I'm trying to restrain a bond length between 2 atoms and a dihedral angle among 4 atoms. I expect tclforces and extrabonds will give close results (or not?) though fluctuation is expected.
> By using extrabond feature I get expected results. But when using tclforces the result is not what is expected. Can anybody shoot me some suggestions please?
Cursory inspection: code looks all right. I can't say more now. You
may need to add verbose debugging output.
> Another question is: from the tclforces example, the potential form seems to be V = 1/2*k*(x-x0)^2 for dihedral, but in the source code of NAMD (also in the manual), for example, in the function computeForce of ComputeBonds.C, the potential seems to be in V = k*(x-x0)^2. Which form should I use in the tclforces?
Any form you like: it's your code! NAMD's convention is inherited from
CHARMM. If you want to call 'k' a harmonic force constant and be
understood by people who don't use CHARMM, then use the 1/2 factor.
If, on the other hand, you see 'k' as a bond stretching constant à la
CHARMM, then don't use 1/2.
This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:21:03 CST