TACOS (Template-based Assembly of Complex Structures) is designed to model the structures of protein-protein complexes based on a hierarchical approach of threading refinement simulations. Starting from a pair of sequences, protein complex templates are first identified from the PDB by multi-chain threading algorithms COTH and SPRING. The fragments excised from the threading template alignments are then used to reassemble the complex structure by replica-exchange Monte Carlo simulations based on a composite force field consisting of knowledge-based interface interaction terms and threading template restraints. Final models are selected based on structure clustering by SPICKER and refined into full-atomic models by ModRefiner (More about the server ... ).
The output of the TACOS server include (An example of the TACOS output):