How to treat dsDNA in COB

[nataled]

COB ‘molecular entity’ definition: "A material entity that consists of two or more atoms that are all connected via covalent bonds such that any atom can be transitively connected with any other atom."

The definition would exclude something like dsDNA. As it is, dsDNA would be considered a complex of molecular entities. This is fine, but it's unclear if that is the intent. We also would have a non-intuitive split between dsDNA and ssDNA (the latter being molecular entity). Again, unclear if this is the intent.

[cmungall]

Thank you. I believe your analysis is correct. We should make this explicit by including dsDNA and ssDNA in cob-examples.owl

I personally agree it's a non-intuitive split, but lets first articulate the example

[dosumis]

Generalise phrasing to cover stable molecular bonds in general? Think this would work to include hydrogen bonds in dsDNA, but exclude transient H-bonds (e.g. between water molecules). There are bound to be edge-cases and judgment calls though.

[nataled]

There is precedent for this. While I can no longer read it, I believe it's been discussed in this opinion article: https://www.chemistryworld.com/opinion/what-is-a-molecule/9391.article

I've seen a number of definitions of molecule, each with different nuances. I'll go through a few that I think might help: 1) Covalent bonding vs chemical bonding is one aspect; the latter includes ionic (but not hydrogen) bonds. Ionic bonds are the sort that hold Na+Cl together. Nonetheless, if more-than-covalent is acceptable, why not allow hydrogen bonding too? I note that the IUPAC definition of molecule invokes the notion of a molecule corresponding to "a depression on the potential energy surface that is deep enough to confine at least one vibrational state." I believe that's code for "stable", and I'd argue that it would allow for any bonding type that allows for stability. If that's the case, then 'stability' will have to be combined with some other aspect of molecule, such as...

2) Determinate vs indeterminate number of atoms. Crystals and lattices would not be molecules since there is no set number of components, similar to the distinction between a complex and an aggregate. I think the determinate number aspect can be useful to us if we decide to allow hydrogen bonding.

Interesting closing sentence in wikipedia: "Whether or not an arrangement of atoms is sufficiently stable to be considered a molecule is inherently an operational definition. Philosophically, therefore, a molecule is not a fundamental entity (in contrast, for instance, to an elementary particle); rather, the concept of a molecule is the chemist's way of making a useful statement about the strengths of atomic-scale interactions in the world that we observe."

Removing the requirement for covalent bonding and instead going with the notions of stability and determinate number of atoms will bring dsDNA into the realm of COB molecular entity and will re-unite it with ssDNA. However, I suspect it will also bring in all of what is currently COB complex of molecular entities. I think here that closing sentence mentioned above comes into play; while I have no problem considering DNA to be a single molecule, I find it difficult to think of some particular protein complex as a single molecule.

[bpeters42]

I see no a problem in having the two strands in double-stranded DNA to be separate molecules. The same way that we consider different protein chains in e.g. the influenza HA trimer to be separate molecules. Covalent bonds are a nice actionable definition how to separate out complexes. Using 'stability' in general will be very hard for users to determine consistently.

I do like the 'determinate number of atoms' - which is a nicer formulation than how we had in the last workshop tried to spell out a distinction between 'molecule' and 'macromolecule'. I am wondering if we can have something like 'determinate molecule' which would be fully defined by an INCHIE string?

[nataled]

From a biologist standpoint, I do find it unintuitive to have dsDNA and ssDNA separated. I imagine most users would as well. As mentioned before, as a biologist, and as you mentioned, I similarly find it unintuitive to think of a protein complex as anything other than being made up of separate molecules of protein chains. For what it's worth, I think someone seeking to argue that a protein complex is a single molecule has a stronger case than someone arguing that dsDNA is composed of separate molecules. I've been trying--so far without success--to figure out what makes me think of dsDNA as a single molecule while not believing the same about protein complexes. The best I can come up with is some appeal to 'natural state' but that seems lacking in rigor.

I'm not fully sure, but my belief is that the determinate vs indeterminate distinction would not help with molecule vs macromolecule, as something like dsDNA is determinate with respect to atoms. The distinction is used to separate molecules from lattices/crystals.

[cmungall]

on the call we discussed that we need to improve this definition:

A complex of two or more molecular entities that are not covalently bound.

should explicitly refer to bonding