Model disagreement with template structure

I have submitted the sequence of pyrin protein to the I-TASSER server. The generated structure was created using the 2wl1A template from PDB, which is one of the domains of pyrin. The latter contains many strands in its secondary structure, but the model generated by the server lacks them. There are helical structures instead. I would like to know the reason and the ways for solving this problem.
Thank you.