Can someone tell me what is the (best) way to extract fitted models after an iterative_structural_align() ?

I have tried something like : 1 aln = alignment( env ); for model in models : aln.append_model( model ) 2 salign.iterative_structural_align( aln ) 3 aln.salign(write_fit=True) 4 aln.write(alignment.ali)

The structural alignment seems OK, but the alignment file alignment.ali doesn't reflect the observed structural alignment. I then realized the salign call probably messes up with the alignment. I thus switched lines 3 and 4, and the resulting alignment seems OK. In fact it resembles what I was expecting according to the structural alignment, and this puzzles me a lot since I somehow expect that after an salign call, the fitted structure and the alignment would be connected.

I also tied something like :

3 aln.write(alignment1.ali) 4 aln.salign(write_fit=True, improve_alignment=False, feature_weights=[0]*6) 5 aln.write(alignment2.ali)

The fitted structures are unchanged but alignment2.ali is totally messed up compared to alignment.ali when I expected the two to be similar. In the end, I am not sure anymore that the fitted models I retrieve are the ones coming out from the salign.iterative_structural_align() or if my salign() call does something in-between...

I'd be glad if someone can give me the best way to retrieve my fitted models. Some hints to explain the observed behavior would be appreciated to.

--Ben