Dan, The easiest solution is to omit that sequence segment from the template alignment which has no corresponding coordinates in the coordinate file, and treat this segment from the view point of target sequence as an insertion. Since there are no coordinates in the template for this region, you do not loose anything by deleting this part of the template sequence. Afterwards, the 10 residue insertion can be modeled with the loop routine. Andras > > Subject: Re: "MODELLER SUPPORT": Re: Adding missing coordinates > Date: Thu, 06 Dec 2001 12:44:08 -0500 > From: Andras Fiser <andras@tuba.rockefeller.edu> > Reply-To: fisera@rockefeller.edu > Organization: The Rockefeller University > To: Dan Thomas Major <majord@mail.biu.ac.il>, nodeller_usage@guitar > References: <3C0F7B7A.91E5DEFC@mail.biu.ac.il> <3C0FA05D.7CAF2074@tuba.rockefeller.edu> <000901c17e7c$ac3232e0$220d4084@biu.ac.il> > > Hi Dan, > > OK, I did not know the exact situation. You need to trick it manually. > > If a whole segment is missing, that is a problem, indeed. You need to add at > least some ad hoc coordinates, e.g. manually to the pdb file. It does not > matter if they does not make any sense (in the loop region) because that > segment will be completely randomized when loop modeling will be done on it, > subsequently after model building. I assume that was your plan. > > But you are right, modeller does require some kind of coordinate entries, for > each residue, at least a backbone trace for the given region as input. > > Andras