need more tether positions on ribbon

[Greg Couch]

When drawing nucleotides representations, ladders and sugar tubes, would like to connect directly to the ribbon, instead of an atom tethered to the ribbon. Ideally that would be where the plane of the base intercepts the ribbon. Chimera1 uses the spline position halfway between the C5' and the O3'. But a functional approximation is the C3' tether position on the ribbon. Would also like the ability to tether a second atom to the ribbon. The second tether would also be useful for proline.

So this is a request for the tether positions of the minimal backbone to be cached.

[Conrad Huang]

Implemented in 2564f1bc1, at least partially.

AtomicStructure instances now have a boolean ribbon_want_backbone attribute which triggers saving ribbon coordinates for backbone atoms. The coordinates are not always saved for efficiency purposes. To get the ribbon coordinates for an atom in an AtomicStructure instance, use structure.ribbon_coords(atom) , which either returns a numpy array of three floats or throws a KeyError exception.

Is this sufficient?

[Greg Couch]

Setting ribbon_want_backbone to True doesn't save the ribbon coordinates. The ribbon needs to be recomputed.

[Greg Couch]

Would like atom.ribbon_coord() to use instead of atom.structure.ribbon_coords(atom).

[Greg Couch]

Implement prototype in nucleotides. Needs structure.ribbon_want_backbone = True to compute ribbon tether positions.

[Conrad Huang]

The backbone positions can only be calculated as part of rebuilding the ribbon graphics. I've changed ribbon_want_backbone from a simple attribute to a property that, when set to True, sets the "ribbon graphics changed" flag so that the graphics will be updated "soon".

Also added Atom.ribbon_coord property to simplify API. The code actually just calls back to the structure for the coordinates but that may be optimized later if necessary.

[pett]

Milestone renamed