show nucleotide representations for modified nucleic acids

[Elaine Meng]

Doesn’t the mmCIF file or some lookup table available from wwPDB say what is the unmodified version of the modified base?

Looks like any modified base (nonstandard nucleic acid residue) is shown as all atoms and ignored by "nucleotides". If there is knowledge of the corresponding unmodified type, consider having "nucleotides" (including initial display) attempt to show it in the same way as would be used for the unmodified base instead. Lots of nonstandard nucleic acids are just hydroxylated and methylated variations of a standard residue. If the required atom(s) aren't there, it could fail back to all atoms for that residue.

Similarly, could also apply the nucleic coloring, or perhaps a slightly modified (e.g. lighter or darker than for the unmodified residue) version of the nucleic coloring.

[Greg Couch]

Looking at 1h0q, the modified nucleic acids, that are based on standard amino acids, are depicted abstractly. But the color isn't changed. How important is that?

[Elaine Meng]

Hi Greg,
The coloring is not that important in my opinion, as people can color these residues individually as they like (if I understand correctly), and the coloring is mainly to indicate the standard residues.

The main thing is to be able to show them as slabs or stubs, at least if they include a similar ring system to their unmodified versions. In the cases I looked at, they weren't, so it would be good to check a few different structures (unless something changed since then). I don't remember the specific PDB IDs, if I didn't include them in the ticket already -- I don't recall the specific PDB ID being important since as far as I could tell none of the modified bases were shown with slabs or stubs.
Thanks,
Elaine

[Greg Couch]

In 1h0q, LKC and LCC are not show as slabs, but TLN, LCG, and LCA are. That is because the chem_comp.mon_nstd_parent_comp_id is ? -- in LCG is DG. LKC and LCC are shown as stubs when switching to the ladder representation because the hbonds are missing from the mmCIF file. If hbond is invoked first, then ladder shows up everywhere. I consider that a bug in the mmCIF from the PDB. It should include h-bonds for all nucleic acid pairs.

[Elaine Meng]

More examples that are not slabbed even though they clearly have a ring system.  Also when the others are tube/slab (default initial display), these are just floating in space because some of their bonds are hidden:

9DG deazaguanine  example: PDB 8omr
ANG aminoguanine  example: PDB 1lwv

[Greg Couch]

In those two examples, the nucleotides that are not slabbed, are not connected to a backbone. It should still be possible to make them into slabs.

The floating in space bug, is a bug that can be worked around by removing the cartoon and reapplying it.

[Elaine Meng]

Oops sorry, bad examples - I thought they were bonded to the backbone but the glycosidic bond was undisplayed - they're almost in the right place for that.  My hasty ligand search apparently only found separate residues, not in-chain ones.  I may look for examples later if/when I actually have time. 

[Greg Couch]

The floating in space bug was fixed in #5732.

Not planning to add support for showing singleton nucleic acids as slabs.