Space group dialog on PDB imports

General Discussion
1

I noticed while I was writing documentation that often importing from the PDB opens the space group selection dialog. I’m not sure how critical it is that the space group information is set, but given it’s not trivial for a beginner or even an intermediate chemist/biochemist to know the space group of a given structure I’m not sure it’s necessarily a great idea to always prompt the user for the space group.

There are two (three?) routes that I can see for improving this.

One is to just disable to prompt for PDB imports, if at all possible. This is probably the simplest (if not easiest) thing to do, and likely won’t cause anyone too much grief since that seemed to be the default for quite a while.

The second is perhaps a better option, but could be problematic vis-à-vis translation/feature freeze, and is to provide an “Unkown” option, since even picking triclinic (P1) implies that more information is known than is in reality. One way around the translation problem might just be to make the x or Cancel buttons select this “Unknown” space group, because then that will stop the dialog from popping up every time you switch back to that document.

The third is probably the most risky, and that’s to provide an “Automatic” space group detection button in that dialog. The button would connect into the Crystal > Space Group > Perceive Space Group command that already exists, and use the default tolerances, hopefully no prompting (it’s supposed to be easy for people who don’t know what they’re doing). This would be pretty useful, but it could just encourage people to always hit “Automatic” and never double check the space group, which is probably not great.

Of all of the options, I think I’m most inclined for now to just go with the second option, and use the x and Cancel buttons to automatically select an “Unknown” space group, since this would require no extra translation and generally means there aren’t any assumptions being made about the structure.

2

Technically speaking PDB files do include the space group in the CRYST1 record: wwPDB Format version 3.3: Crystallographic and Coordinate Transformation Section

It would be a bit tricky to keep the unit cell information for PDB and prevent the space group choice.

My suggestion would be that most people importing a PDB file don’t care about the unit cell and we can ignore it.

3

I’m assuming that not all structures have that information, and not that Avo2 is just ignoring it?

I would support this, although I am not a biochemist so my opinion probably isn’t worth much.

4
5

That’s correct. Keep in mind, while small molecule crystallographic models typically are stored in .cif (link to IUCr’s documentation), .pdb are preferred to document structures of large (bio)molecules.

One way to the .pdb is to obtain and successfully examine crystals of your protein by X-ray, neutron, or (cryo-)electron diffraction; which however can fail for multiple reasons. Another approach, sometimes actually wanted because it keeps the proteins in a state more flexible (and thus more similar to what happens e.g., in a cell) is to record 3D NMR spectra / triple resonance experiments in solution, a technique popularized for instance by Kurt Wüthrich. These data are equally part of PDB (the database), see here. While I don’t know how many ssNMR based .pdb are deposited in the PDB, in solution space group symmetries tend to be lost.

A .pdb derived from an experiment carries keyword EXPDATA to inform the user accordingly (see pdb’s documentation about the title section).

6

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.