some dihedral angle libraries has improved, but in case of significant sequence similarity you will not see much effect of it.

One difference, that the reference to the refinement protocoll has changed, earlier refine1 was the most exhaustive and refine 3 is the least, now it has changed and refine 3 is more exhaustive than refine2, and refine2 is more exhaustive than refine 1.

Andras

> -----Original Message----- > From: owner-modeller_usage@salilab.org > [mailto:owner-modeller_usage@salilab.org]On Behalf Of Luca > Settimo > Sent: Thursday, January 31, 2002 9:04 AM > To: 'Modeller Usage List' > Subject: difference between version 4 and 6 > > > Dear modellers, dear Andej Sali, > > I have one question for you. > I built several protein models using modeller 4 time ago... > I used in the top file the following parameters: > > SET HETATM_IO = on and off > SET WATER_IO = on and off > SET HYDROGEN_IO = off > SET STARTING_MODEL = 1 > SET ENDING_MODEL = 1 > SET MD_LEVEL = 'refine3' > CALL ROUTINE = 'model' > and I build the model with and without ligand and water in the binding > site... > > since the models that I built are quite a lot my question is: > > > do I have to expect some difference with the models obtained using the > new version (modeller 6 released the 25 January 2002) ... or do I have > to expect the same result? > for example does something differ when I model the ligand in the two > version? or just by using the same parameter in the top file above here? > > > > Thanks > > Luca > > ................................................................... > > Luca Settimo > Molecular Structures and Biocomputing > Department of Biochemistry and Pharmacy > Åbo Akademi University Tykistökatu 6A 20520 Turku Finland > phone: +358 (0) 2 2154600 > fax: +358 (0) 2 2153280 > mobile phone: +358 50 5238418 > mailto:lsettimo@abo.fi > http://www.abo.fi/~lsettimo > > >