Weird looking molecules

22 July 2023 - Graphics

Weird week with weird molecules. This site makes use of SmilesDrawer for rendering molecules and reactions directly from a SMILES string with Javascript in the browser. This library works fantastically with only two known shortcomings: it does not know what to do with some trans double bonds and it confuses Mesyl groups (Ms) for acetyl groups (Ac). The last issue I was able to solve (pull request pending) and problem one magically disappeared by simply converting the trans bonds to undefined bonds with a lot of them coming out as trans anyway.

Two known issues until this week, when patrolling the reaction database, I stumbled on this weird molecule:

which is supposed to represent the compound 7,7-difluoro-1-((4-methoxybenzyloxy)methyl)bicyclo[4.1.0]heptane. What went wrong here? The same official SMILES string FC1(C2CCCCC12COCC1=CC=C(C=C1)OC)F renders perfectly fine in the Indigo toolkit as used on the Opsin webpage. It turns out the SMILES string can be tweaked in a way that the structure is again presentable: C3=CC(OC)=CC=C3OCOCC12CCCCC2C1(F)F.
Mind you , this is the first molecule misbehaving in SmilesDrawer out of a million molecules in the CRD database and if somewhere in the future the javascript code gets a fix, the messy molecule listed first (bonds wildly going all over the place) will again look identical to the second one. The second weird molecule detected this week was spotted while patrolling the journal Organic Letters for useful synthetic procedures. One publication apparently announced the synthesis of 4,4,4,4,4,4,4-heptafluoro-1-(2-methyl-1H-indol-3-yl)-4λ8-but-2-yn-1-one, a pretty wild claim because it looks like this:
Notice the methyl group with 7 fluorine substituents. Twitter was quick to offer an explanation: Chemdraw has an image to systematic name conversion tool but due to a known bug all perfluoroalkyl chains end up like this. An overworked PhD student must have then pasted it in the supporting information, no need to blame anyone. You may ask why complain on Twitter and not read the accompanying article first? Well, the focus is my work is collecting organic procedures which is already a time-consuming procedure as it is. Ideally the procedures in the supporting information are self-contained without needed any context. Anyway, this is the cleaned up version as it will appear in the CRD database:
The article is an Organic Letters ASAP publication, the author instructions are abundant but few words are spent on the peer-review process. Are ASAP articles in the queue for peer-review? Does it include the supporting information? Answers please, ASAP!