# [gmx-users] manual eq. 4.74-4.75 (dihedral restraints) head scratcher

Daniel L. Ensign densign at mail.utexas.edu
Sat Apr 17 17:29:53 CEST 2010

Hello gmx-users, you rock and rollers,

Equations 4.74 and 4.75 in my copy of the manual have (please pardon
my pseudo-LaTeX):

(4.74)
\phi' = (\phi - \phi_0) MOD 2\pi

(4.75)
V(\phi') = \frac{1}{2}k ( \phi' - \phi_0 - \Delta \phi )^2 if \phi' >
\Delta \phi
or
V(\phi') = 0 if \phi' \leq \Delta \phi

but should there be absolute values around all of the \phi-\phi_0?
Which way is it in the code -- with the absolute distance between \phi
and \phi_0 or the directed distance?

Also, as far as I can tell (and some mornings I definitely don't read
too good) neither the manual nor
http://wiki.gromacs.org/index.php/Dihedral_Restraints define the
fields in [ dihedral_restraints ], although the latter does name them.
There, I see

[ dihedral_restraints ]
; ai   aj    ak    al  type  label  phi  dphi  kfac  power
5    7     9    15     1      1  180     0     1      2

ai, aj, ak, al = atom numbers, obviously
type = ?, but I'm guessing there's only one type anyway
label = what is this one?
phi means phi0 ?
dphi means Delta phi, I guess
kfac is the force constant, probably
power = what is this one? Does 2 give me harmonic constraints?

I'd be happy to translate any answers given to the wiki, assuming I
get answered and that I'm allowed to edit the wiki.

May the Force be with you (as long as it's calculated in hand-tuned
assembly loops),
Dan

--
"Quas Exact Comprehendere et ponderare Dei est non hominum." - Juan de Salas