This server module predicts the atomic level structure of a protein complex. Given at least two protein sequences, this module first searches for homologs in a database of subunit sequences of experimentally determined protein complex structures. If all input sequences are homologous to subunits of an experimentally determined complex, our multimeric comparative modeling will be used to build three dimensional protein complex models for the sequences. Results are sent out by email in PDB format. (In some cases, several models are predicted and the best five are returned.) This service provides a user friendly way for researchers to create all atom models of protein complexes of interest. The all atom models can provide a useful insights into the functional and mechanistic details of the protein complexes as well as information on the residues that mediate protein complex interactions.
The protein complex library contains 20,588 protein complex templates. New users may want to look at this short tutorial which includes some notes on how different mailers handle returned data.