Make initial color of molecular surfaces match ribbons

[Tom Goddard]

Currently the initial color of molecular surfaces uses color by chain (for per-chain surfaces). But many atomic models are initially shown using color by polymer, so the surface colors don't match the initial atomic model color. Also some atomic models (4 or fewer chains) show all chains as the same color but new surfaces have different colors for each chain.

I think it would be better if molecular surfaces had initial colors that matched the atomic model colors. Often the atoms of chain are not all the same color, for instance ligands are often colored byhet. But the ribbons are usually given a single color per-chain, so having the surface color match ribbon colors might be best.

Currently surfaces start with a single color. What if the residues of a ribbon have different colors? Should the surface be multicolored? I think it should still be single color. In that case we might use the average ribbon color of the residues.

[Elaine Meng]

Seems like if you use "color fromribbons"  it should give color patches on a molecular surface if different residues have different ribbon colors.  If a different method is used to give initial surface color, then people will want to have a command for that too.    Usually the ribbons will be a single color, however, except when people color the atoms of certain residues and haven't figured out how to do that without also changing the ribbon color.

[Eric Pettersen]

And is "fromribbons" the correct term (vs. "fromcartoons")? Which term do we use in commands (or support both?)?

[Elaine Meng]

I would favor "fromcartoons" or both (fromcartoons, fromribbons) -- I believe that we generally support both, but I give "cartoon" top billing because the tube helices aren't ribbons.

[goddard@…]

Ok I've made both fromcartoons and fromribbons do the same thing.

[goddard@…]

Yes "color fromribbons" will make a multicolor surface if the residues have different ribbon colors, while the initial surface display I suggest should be a single color.  Currently surfaces start with a single color.  Gaussian surfaces (with resolution specified) don't have atom associations so can't be easily colored by residue.  I'm not sure there will be much demand for coloring the surface using this initial surface rule and in most cases "color fromribbons" will do that.  So I don't foresee the need to add another color command to color surfaces by average ribbon color.

[Tom Goddard]

Done. In 1.6, not in 1.5.

Made the initial molecular surface color be the average of the cartoon colors for the residues defining the surface. This seems like a big improvement.