I have the following question: I wish to model a series of homologous proteins (28 in total) for which I have the structure of the first and the 26th in order of sequence homology (ranging from ca. 55% to ca. 20%). The protein is made of two domains. The two available structures are quite similar except for the orientation of one domain with respect to the other (significantly off). What is the best approach to model sequences #2 to #25 and #27-#28??? Is there an accepted protocol in order to model multidomain proteins? I know that one of the biggest problems with modelling protein structures is encountered when you have more than one domain, that is, to model relative domain orientation. Is there a way out? How would you proceed using Modeller???