Coordination clash when constructing a dimer structure with loop missing

Dear everyone

The attached pdb file (dimer1.pdb,https://www.dropbox.com/s/dkvg17bgp9sun5n/dimer1.pdb?dl=0 ) is a dimer structure containing two same monemers. In each monomer, a loop (residue 535-572) is missing (please see the attached complete protein sequence file: full sequence.fasta,https://www.dropbox.com/s/0lrc7v3gptyiuas/full%20sequence.fasta?dl=0). Because the gaps on the two monomers are very close to each other, when using I-TASSER server to model the chains individually and then use structure alignment to construct the dimer, the loops added by I-TASSER server clash into each other (attched file: dimer_loop_filled.pdb, https://www.dropbox.com/s/6j7avzpvu094czg/dimer_loop_filled.pdb?dl=0).Even though the loop filled dimer was refined by FG-MD, the problem became more serious because the loop added on one monomer clashed into the other monomer (attached file:dimer_loop_filled_refined,https://www.dropbox.com/s/ik7p9loqbq6n3t4/dimer_loop_filled_refined.pdb?dl=0 ). I hope the added missing loop should leave as far as possible away from each other and from the rest part of the remain structure. Could you give me some tutorial on this task by using I-TASSER locally?

Professor Yang Zhang suggested me to download I-TASSER and manually construct the loop regions with the remaining part frozen.However, I cannot find such a usage in anyI-TASSER documents. Could you give me some tutorial on how to do this?

Thanks in advance.

Yeping