I’ve started a review of the DailyMed dataset as it is representative of some of the struggles with preparing a curated dataset of chemical structures, chemical names and trade names. In the first comment I pointed to issues with structure representations. I believe one of the worst is shown for qvar to the left. An examination of the qvar record gives the name as beclamethasone propionate. This particular compound has the chemical structure shown below. Not only is the stereochemistry missing from the structure on DailyMed but also half the ring has been lost, maybe during a scanning process? I wonder whether the label circulating out there to the public has this issue? Would the public care? Probably not. But when trying to build a curated dataset it’s rather important.


Stumble it!

One Response to “Daily Med Commentary 2 – More Bad Structure Representations”

  1. Chris Singleton says:

    I agree that the public may not be interested, but this is pretty egregious. I couldn’t tell that that first structure was a steroid until I saw the correct structure below. Frankly, if it’s worth doing, it’s worth doing correctly, so this just shows the importance of curation.

Leave a Reply