By visiting http://autodock.scripps.edu/wiki/Introduction, one can read:

"What's New in AutoDock 4

AutoDock 4 offers many new features and improvements over previous versions. The most significant is that it models flexible sidechains in the protein. We have recalibrated the free energy scoring function using a much larger training set of ligand-protein complexes where we know both the 3D structure and the inhibition constants, Ki. There are also new search methods, thanks to our ongoing collaboration with colleagues at UCSD and Sandia National Laboratories.

Many limitations of version 3 have been removed in AutoDock 4. One of biggest annoyances was the limit of 6 different atom types in a ligand; this has been increased to 22, and the method of specifying atom types is also improved, using a single parameter file called "AD4_parameters.dat" by default. Also, AutoDock 4 no longer requires the user to set the stacksize to unlimited in the UNIX shell."

First of all, I beg your pardon whether I feel such an ignorant about the whole chemical subject and might be picky and boring to some people.

Ok, presuming AutoDock 3 and W.C.G. agent are the same sofware, I came upon an idea (useful for crunchers who won't understand much further, like me):

If AutoDock4 would be better than its predecessor, AutoDock 3, why is it (AutoDock4) so complicated to be running on a PC?

Has anyone thought about this as a means of improving AutoDock4 up to the point of making it as "automatic" as W.C.G. agent FightAIDS@Home?


Regards,

Dan60