Quoting John Chodera <jchodera.gmail.com>:
> I was hoping to find someone who had already done so to spare us the
> effort, but we may have to do this ourselves after all, and this
> information will come in handy. :)
My feeling is that those new residues are not available for the Duan
et al FF... You will have to do it ;-) The problem with the Duan et
al. FF is that the theory level used to compute the MEP is quite big &
the corresponding cpu time can be important for complex projects.
Without wanting to upset you, I would like to comment "to spare us the
effort" in the AMBER mailing list as many users have the same "reflex"
;-)
- For new residues, using FF parameters & charges from others is
convenient. However, developing your own set of parameters/charges is
very informative. You study the QM potential energy surface of your
new groupement(s) (i.e. that you will have to match during MD
simulations). So the effort you pay at the beginning will allow you to
better understand/check your MD simulations.
- In the original Kollman strategy, the role of RESP charges was
central. The idea was to have a "simple" FF with a robust (& complex)
charge model (vdW, electrostatic & dihedrals are the three major
components for a good FF). Moreover, charge values are used to derive
FF parameters (& among them dihedral FF parameters). The same idea is
still used in the last AMBER & GLYCAM FFs where FF developers spend
time for deriving "good sets of atom charges".
In parallel, people (mainly new users) are interested in deriving
charges "on the fly" for AMBER FF. My personal opinion is that these
charges are not "RESP charge values as originally defined by the
Kollman's group". However, here, I might not be the best person to
talk about that...
regards, Francois
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sat Jan 05 2008 - 00:03:37 PST
