Dear Modeller users,

To study the interface between two proteins in complex (A+B), i.e. to decide whether these proteins may form a complex, I wish to compute a homology model on the basis of an existing, but only remotely homologous, complex (A'+B'), i.e. there are difference in some indels close to the interface.

Each of the target proteins has also a closely homologous template, (A" and B"), which I would like to use to model the indels.

However, as I am particularly interested in the interface, it is crucial to model the sidechains at the interface correctly, that is to account for the presence of the binding partner. Thus, the remote homologue (A+B) bears some important information as what concerns the complex.

The most direct approach would be to use all templates and let MODELLER decide what to do. But is this really the optimal approach?

Thank you very much in advance for any suggestion, or reference to similar problems.

Kind regards,

Karsten.