Issue #22 resolved

How to deal with Deuterium?

created an issue

Hi Dan,

Does OPSIN handle deuterated compounds? If if does, how would it expect these names to be formatted?

Thanks, Paul

Comments (4)

  1. Daniel Lowe repo owner

    OPSIN supports deuterated compounds using the prefix deuterio (alternatively deutero) as a substituent indicated replacement of hydrogen by deuterium e.g. (S)-1-deuterio-ethanol [tritiated compounds can be done using tritio]

    The IUPAC defines a far more general way of doing isotopic replacement which is supported by Marvin e.g. (1S)-(1-²H₁)ethan-1-ol [On at least the version I have, Marvin can't read this back in which I expect is why you are asking ;-)]

    Supporting this method is definitely in scope for OPSIN but it hasn't been a very high priority as its use seems to be relatively rare in practice

  2. pmortenson reporter

    Yes, the Marvin-style output is what we have. Basically when generating names for e.g. patents, we use Chemaxon to generate the name, and then run it through OPSIN as a check. We do have some deuterated compounds, which Chemaxon outputs in the more general way you describe above.

    We actually don't have the Chemaxon name-to-structure package - it was quite a lot more expensive than structure-to-name, and in our tests it didn't perform as well as OPSIN - even on their generated names! Additionally, I like the idea of using a different package as a more independent check on the name.

    Thanks for the response - I can appreciate that this is probably not a common problem!

  3. Daniel Lowe repo owner

    Hi Paul, I've just released a new version of OPSIN that adds supports for this. As well as the IUPAC style e.g. (1S)-(1-2H1)ethan-1-ol. The style used by CAS/ChemDraw generated names is supported e.g. (S)-ethan-1-d-1-ol. Please let me know if you encounter any issues.

    (...I've also just noticed that I've made a slight oversight of forgetting to add the conversion of the Unicode subscript characters to their ASCII equivalent e.g. ₁ -->1.... which I'll fix now but isn't in version 2.2.0)

  4. Log in to comment