> > Or are you saying you˘d like something like this put into the standard
> > pipelines?
>
> Exactly - something like this would be a useful (default) option in the
> MR packages. I think with the current computational power it is quite
> doable.
And in this way we can disguise a failure in proper training as a pitfall
of the automated MR computer programs. But if we are to go this way, the
easiest solution would be to make the programs completely black-box (with
closed source) and re-program them to return a 'sorry, try again, no
convincing solution found' instead of returing the best solution they
located (which would allow the untrained individual to push ahead and
waste time and effort).
--
Nicholas M. Glykos, Department of Molecular Biology
and Genetics, Democritus University of Thrace, University Campus,
Dragana, 68100 Alexandroupolis, Greece, Tel/Fax (office) +302551030620,
Ext.77620, Tel (lab) +302551030615, http://utopia.duth.gr/glykos/
|