"sym" copy new model changes cartoon from original

[Elaine Meng]

Basically 2 helices that are adjacent but shown as separate helices in the original structure are shown as one continuous helix in the sym new model, and the latter makes for bad tubes. Example is 5j7a and the juction between the two helices is /a:163,164 colored red in the attached images. The images show the monomer original structure and the trimer sym copy created with "sym #1 assembly 1" (or clicking link "1" in assembly table in Log).

Don't know where the problem lies (data representation? data copying?) so CCing Conrad.

[goddard@…]

Likely a problem where the structure copy is not copying something.  I'll investigate.

[Eric Pettersen]

There’s a large Python-layer component to cartoon depiction.  If you look at the Structure session-saving code you can get an inkling of what’s involved. The C++ layer attributes are properly copied.  Maybe you need to enlist Conrad in implementing the Python-layer cartoon attribute copying (there’s a class instance involved…).

—Eric

[Tom Goddard]

The problem is that AtomicStructure.copy() does not copy the Residue.ss_id values (all values are -1 in the copy). So the copy does not realize there are two adjacent helices with no residues in between. Probably this is just an omission. I'll let Eric fix it since he wrote the copy code.