From: Ross Walker <>
Date: Fri, 12 Dec 2008 09:06:22 -0800

Hi Saurabh


To follow up on Carlos's excellent explanation I would just like to add that
this of course only applies to the use of wildcards (X-) in dihedral terms.
This is something that has always concerned me in AMBER because it makes it
very difficult to understand what is going on. Additionally the assumption
(made by the idivf value) is in my opinion dangerous since one could almost
certainly find molecules (not necessarily in the amino acid set but
elsewhere) that had different connectivity for that wild card dihedral.


My advice if you are generating your own parameters is to ALWAYS explicitly
define all dihedrals and in all cases set idivf=1 and set the barrier height
appropriately for that specific dihedral. This will avoid much confusion and
chances for mistakes.


All the best



From: [] On Behalf Of
Carlos Simmerling
Sent: Friday, December 12, 2008 6:05 AM
Subject: Re: AMBER: ff99 (PK/IDIVF)


the k value used in the dihedral function is pk, but this is the total for
rotating that bond (in this case 14.5). since there are multiple sets of 4
atoms that contribute, the total is divided by the idivf, the expected
number of 4-atom-sets that will have this applied. each will get in this
case 1/4 of the total (1 over the first number after the x-c-c-x in the .dat
file), or 3.625.
the 2 in red is the periodicity of this dihedral term - see

On Fri, Dec 12, 2008 at 8:26 AM, saurabh agrawal <>

Dear Amber Users,

In xleap leaprc.ff99 parameters looks like:
                 N Pk/IDIVF Po
?-c-c-? 2 3.6250 Pi

While In parm99.dat file torsional angle parameters are as given:

x-c-c-x 4 14.50 180 2

May some one explain the meaning of Pk/IDIVF (red) field in torsional angle
parameters in the ff99 force field and and the formula to calculate it.
Also the caption of the fields in the parm99.dat file.

Sincere regards,

Thanking you in advance.




Add more friends to your messenger and enjoy! Invite them now.


The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
Received on Sun Dec 14 2008 - 01:11:04 PST
Custom Search