Changes between Initial Version and Version 2 of Ticket #5386

Timestamp:

Apr 3, 2026, 2:15:39 PM (6 days ago)

Author:

Eric Pettersen

Comment:

This is theoretically fixable, but would be significant work. The problem is that because of the forced association of the structure sequence with a different sequence, there is an effective deletion in the structure relative to the sequence. The code generating the Modeller input doesn't expect/handle this. The solution would be to put a gap into the target sequence so that the template sequence could then include the skipped structure residue(s), but also the indices into the sequence indicating the remodeled region might have to be adjusted. The code is already pretty complicated, and it would be delicate work to make these changes. Since this problem has only been reported once in 4 years, I'm not going to try to make these changes unless more reports come in.