Recognize the equivalence of circularity and crosslinks which generate circular molecules.

[paulflang]

The circular attribute seems redundant to crosslinking (at least for proteins). Not sure if we are OK with providing two ways to do the same thing.

[jonrkarr]

Semantically, I think its useful to have a special syntactic sugar for circularity. Semantically, this connotes different meaning than a crosslink.

More generally, BpForms is not founded on the principle that there should only be 1 unique way to describe a polymer.

• For circular polymers, there isn't necessarily a well-defined start position
• There's nothing that prevents the community from defining monomeric forms which are themselves polymers. If such a monomeric form is semantically meaningful on its own and has a commonly accepted id or name, I don't see a problem with this.

The main goal of BpForms is to be concrete. I don't think this requires a unique representation of each molecule.

[paulflang]

OK, thanks. This all makes sense.

For circular polymers, there isn't necessarily a well-defined start position

I noticed that this a problem where the bpforms equality check fails. I had a brief look at the code, but the way it is structured I couldn't find a trivial fix.

[jonrkarr]

Yes, that a good point. We should change the code to recognize the equivalence of this special case. This comes up in a couple places

• equality
• it shouldn't be possible to describe a circular molecule with a crosslink between the same atoms; this should result in an error.