Dear Doug,
AutoDock Vina uses a simpler scoring function than AutoDock. It does
not use atomic charges but rather only atom types, using a different
atom typing scheme internally. The scoring function is based on a
machine learning approach trained against both conformation and
affinity. The gains in performance are, in large part, due to the use
of the gradient in the optimization method that gives it a "sense of
direction" from a single evaluation.
Currently, our theory regarding its improved docking results speaks to
the possibly too strong assumptions that are made in the physics-based
scoring methods, and to the fact that our scoring approach uses both
experimental structure and affinity in optimizing the form and weights
of the scoring terms.
Let us know how it works for you.
Best regards,
Art Olson
Douglas Houston wrote:
> Dear Dr. Olson,
>
> Your lab's docking program sounds very interesting. Since I can't see
> the paper yet, do you have any theories about why your program appears
> to produce superior dockings despite running more quickly?
>
> I presume that using ADT means you merge non-polar hydrogens? Could
> you also tell me how you add hydrogens and assign partial charges to
> your receptor molecule?
>
> I'll definitely give your beta a whirl and see how it goes,
>
> cheers,
>
> Doug
>
>
> Quoting arthur olson <
[hidden email]>:
>
>> Dear AutoDock Users,
>>
>> I want to alert AutoDock users to a new beta version of Vina, which is
>> now called AutoDock Vina, a new docking code developed by Oleg Trott in
>> my lab. AutoDock Vena uses the same format for the ligand and receptor
>> files as AutoDock4, which can be prepared by AutoDock Tools, however
>> its input is much simpler, and there is no need to run AutoGrid before
>> runing the docking. The scoring and optimization procedures in
>> AutoDock Vina are quite different than those in AutoDock4.
>>
>> A paper announcing and describing AutoDock Vina is about to be
>> submitted for publicataion. In the paper we compare its performance
>> relative to AutoDock4. We have found AutoDock Vina is considerably
>> faster than AutoDock4, with comparable or better docked structures for
>> many complexes.
>>
>> The newly released AutoDock Vina 1.0 beta 03 has the following changes
>> from the previous release:
>>
>> (1) Improved estimation of the statistical influence of the
>> flexibility of the ligand on its free energy of binding
>> (2) Optimization of multiple interacting ligands is disabled
>> If you run Vina now with 0 or 2 or more ligands, you'll get an error,
>> saying
>> that this version requires exactly one ligand.
>>
>> If people have a particular need to dock multiple ligands
>> simultaneously, please let us know and we can discuss future
>> possibilities.
>>
>> AutoDock Vina 1.0 beta 03 is distributed as binary executables for both
>> Linux and Windows platforms (sorry, no Mac OSX version yet). It can be
>> downloaded from vina.scripps.edu or via links from
>> autodock.scripps.edu. I encourage you to try the beta version and
>> provide feedback to us.
>>
>> Thanks,
>>
>> Art Olson
>>
>>
>> --
>> @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
>>
>> Arthur J. Olson, Ph.D.
>> Professor
>> Department of Molecular Biology
>> The Scripps Research Institute
>> La Jolla, CA 92037
>> tel. (858) 784-9702 fax. (858) 784-2860
>>
http://www.scripps.edu/pub/olson-web>>
>> ________________________________________________
>> --- ADL: AutoDock List ---
http://autodock.scripps.edu/mailing_list ---
>
>
>
>
> _____________________________________________
> Dr. Douglas R. Houston
> Room 3.23
> Institute of Structural and Molecular Biology
> Michael Swann Building
> King's Buildings
> University of Edinburgh
> Edinburgh, EH9 3JR, UK
> Tel. 0131 650 7358
>
--
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
Arthur J. Olson, Ph.D.
Professor
Department of Molecular Biology
The Scripps Research Institute
La Jolla, CA 92037
tel. (858) 784-9702
fax. (858) 784-2860
http://www.scripps.edu/pub/olson-web________________________________________________
--- ADL: AutoDock List ---
http://autodock.scripps.edu/mailing_list ---