Robert Jorissen wrote:

> Dear fellow Modeller users > > I am using Modeller 4 to build a comparative model and am at the stage > of constructing loops. > Do I accept the loop with the lowest score (as long as it is plausible) > or do I use other > criteria as well e.g. combination of low score and favourable ProsaII > plot over the region of the > loop. > > Rob > > -- > ----------------------------------------------------- > Robert Jorissen > > Ludwig Institute for Cancer Research > P.O. BOX 2008, Royal Melbourne Hospital > Parkville, Victoria 3050, Australia > > phone: ++61-3-9341-3155 > fax: ++61-3-9341-1938 > > e-mail: Robert.Jorissen@ludwig.edu.au > -----------------------------------------------------

Modeller itself optimises the objective function, so you need to select the model/loop with the lowest objective function score. If you generate several models and they have significantly different conformation and/or similar final objective function score than it could be useful to invoke external tests like Prosa, Procheck etc. to make your decision about the selected model more confident. That is not unusual at all, if not the lowest objective function score belongs to the best model, but usually one of the lowest scores belongs to the best available model and it is often the lowest one.

Andras

-- , Andras Fiser, PhD # phone: (212) 327 7206 The Rockefeller University # fax: (212) 327 7540 Box 270, 1230 York Avenue # e-mail:fisera@rockvax.rockefeller.edu New York, NY 10021-6399, USA # http://salilab.org/~andras