Jose,
My group at Merck Frosst has used amber6 and, more recently, amber7
extensively for a number of years for production-level MD in the context of
therapeutic programs. We developed the AM1-BCC charge model to make possible
the necessary higher throughput of charging small-molecules; this is what we
use almost exclusively, and it is implemented in ANTECHAMBER. For the other
parameters, we have a huge frcmod.smallmol file based on the parm94
forcefield; this has been built up over about eight years and so we haven't
tried ANTECHAMBER's small-molecule forcefield; the cost to validate it in
order to switch over is currently too high, although ultimately we will have
to update our small-mol forcefield. The small-mol forcefield, its accuracy
and generality, concerns me deeply and I think it is the Achille's heel for
using amber in the pharmaceutical industry.
The other key issue is the human time required to amber-ify ligands
(especially the atom typing)and building the lig-enzyme complex in leap.
While ANTECHAMBER was made to do some of this, we had to automate all of it
back in the amber5 days, which we did using the Merck & Co. programmable
chemistry perceiver and the inevitable multitude of scripts. We have
continued using our homemade protocol instead of ANTECHAMBER because "it
ain't broke": it works, it is validated (in our hands), and it allows us to
go from model-built starting structure to submitted amber jobs in about two
minutes, with most of that time being the AM1 optimization.
So, "yes" to using amber7 in an industrial environment to test many small
organics against a given target, "yes" to the AM1-BCC charges as being
suitable plug-and-play replacements for RESP charges, and "I don't know" to
the suitability of ANTECHAMBER's other forcefield parameters... we use
parm94-based supplemental parameters.
Christopher Bayly
-----Original Message-----
From: Jose Gallego [mailto:Jose.Gallego.medivir.com]
Sent: October 27, 2003 8:18 AM
To: amber.scripps.edu
Subject: AMBER: amber 7 for industrial application
Hello,
I work for a pharmaceutical company and we are considering purchasing AMBER
7. I have been an AMBER user in previous academic posts, and one frequent
problem during that time was the required parameterization of small organic
molecules into the AMBER force field.
I was pleased to see that AMBER 7 contains the ANTECHAMBER module apparently
capable of automatically incorporating most small molecules into the AMBER
force field.
Since the purchase of the program is a significant investment and AMBER does
not provide demo versions, I would like to receive opinions from people who
have used amber 7 in an industrial (pharmaceutical) environment when one has
to test many small organic molecules against a given target. How does the
program perform for this type of task?
Also, I have read some messages mentioning amber 8. How soon is the new
version likely to be released??
Thanks!
Jose Gallego
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Oct 29 2003 - 16:53:02 PST
