1gmk cartoon strand twist

[Elaine Meng]

Gramicidin has some D-amino acids and in one place the ChimeraX ribbon has a seemingly unnecessary twist, viz /c:7-8 (image attached). The chain A/B ribbons behave OK, but there is this one twist in the C/D assembly. It may be a hard case for ChimeraX (prioritize as desired), but in Chimera all the ribbons are pretty, similar to what is shown in the RCSB page for this structure
​http://www.rcsb.org/pdb/explore/explore.do?structureId=1gmk

open 1gmk
trans 50 targ c
car supp f
hide; show backbone
sel /c:7-8

[Elaine Meng]

Upon closer examination, I don't think this problem with 1gmk is from assuming alternation of carbonyls from one edge of the strand to another. If that assumption was made, this ribbon would not have the wrong twist. This may be a failure of the logic to account for bulges where there isn't a strict alternation.

I don't think this problem is so common, in that case. It looks like you made the algorithm smart enough to allow for the common bulge case already. For example, 2gbp strand :172-180 has a bulge where the 175 and 176 carbonyls are on the same side of the strand, and the ribbon does not twist, so it is being handled correctly. I'll attach an image showing this.

If this new assessment is correct I think we can downgrade this ticket as a rare case. If it is easy to fix, OK, but otherwise not worth the sweat.

[Eric Pettersen]

Milestone renamed

[Elaine Meng]

This bug is apparently fixed, regardless of "wiggle" setting. The problem of the twist as highlighted in the image from over a year ago is no longer present with wiggle true or wiggle false.
​https://plato.cgl.ucsf.edu/trac/ChimeraX/attachment/ticket/676/1gmk-twist.png

I'll investigate the effect of wiggle on additional structures get back to you. Initial impression is that wiggle-true is better, but needs more thorough confirmation.