Hello,
Use all the three templates simultaneously.
modeller will automatically down-weight the information in the overlapping regions (i.e. inside the domains) if it comes from a more distantly related protein. however in the connecting region (i.e. domain orientation as defined by non-bonded contacts) the info will come exclusively from the complex structure template, because the individual domains does not have any.
Consequently the simplest and yet the most effective approach is to combine all three templates in one multiple input alignment.
If you are very worried about the distantly related complex template, because, let us say its conformation within the domains is really different from the others and not credible, you can artificially make its contribution to the restraints even less, by eliminating most of its structure and just keeping a few residues long segment that occurs at the "border" of the two domains, to ensure the proper orientation.
best wishes, Andras
Satyan Sharma wrote: > > Hi, > I have a protein with two domainswhich i would like to model. I have 3 > template structures. One template (T1) has both the domains but with not > good homology, which I would like to use for the orientation of the 2 > domains. The 2 structures are of 2 individual domains with very high > homology. I have tried to model my protein using template 1 (T1)and use > the ".ini" file from it to further refine the 2 domains. Now is there a way > how can i put these 2 back. Any suggestions. Or is there a better > approach. > Thanx, > Satyan Sharma