Computer software for the generation of systematic names from structures and the conversion of chemical names to structures has been the subject of numerous discussions on this blog as we have discussed Chemmantis and the need for high qualityy conversion of chemical names to structures. There are a number of software programs available for the generation of names from structures and vice versa. The ones I have most knowledge of are those of ACD/Labs, Cambridgesoft and OpenEye. I have worked with all of them and they all have their strengths and weaknesses but all three companies are working to improve their products. What is interesting is Cambridegsoft’s position in regards to Name to Structure conversion…they seem to have it patented.

The chemical names that you see on ChemSpider are generated using OpenEye software as shown below. We owe them a debt of gratitude for providing us the software.

openeye

The nomenclature “guru” at OpenEye is Roger Sayle and Steven Bachrach’s post here pointed me to Roger’s recent publication “Foreign Language Translation of Chemical Nomenclature by Computer”. While at ACD/Labs we included both French and German nomenclatre support for generating chemical names from structures. So, I have a real appreciation for the issues of multilingual nomenclature. In fact, when I process files myself for deposition in ChemSpider I do use their desktop tool for generating French and German names so don’t be surprised to see accents in some of the identifiers on ChemSpider.

Roger’s paper is an amusing and educational read and I recommend it to anyone interested in the complexities of nomenclature. He discusses multilingual support including Chinese and Japanese and even wanders into Klingon! OpenEye did the right thing in making this Authors Choice so the pDF is available to everyone here: http://pubs.acs.org/doi/abs/10.1021/ci800243w. Clearly OpenEye have a great path forward for nomenclature and it’s going to be interesting to watch their product develop.

An additional comment…we have to deal with a lot of complex issues regarding fonts on web-based articles. It’s not easy but we do proofread everything on ChemMantis and the ChemSpider Journal of Chemistry and try to catch all the issues. Take a look at the HTML version of the article here: http://pubs.acs.org/doi/full/10.1021/ci800243w. Oh those exploding fonts!

chinese

Stumble it!

2 Responses to “Foreign Language Translation of Chemical Nomenclature by Computer”

  1. Egon Willighagen says:

    What has been annoying me for a very long time is the way chem DB systems deal translations. When the paper was out just shortly (thanx to RSS), I scanned over the text, and did not find a clear solution.

    What a proper chem name translation should know is that:

    “acetic acid”@en translates to “azijnzuur”@nl, and
    “ethanoic acid”@en translates to “ethaanzuur”

    Even though the both refer to the same compound. The ‘Woordenboek Organische Chemie’ did this (the data is available from sf.net/projects/woc), but was the only one to do the translations maintaining ethymology.

  2. Tore Eriksson says:

    Interesting article, but it could have been better if the author had a linguist check it as well. The “Historical influence” part contains statements like:

    “It is interesting to observe that both Chinese and Japanese use the same word (character) for water, ‘水’, predating their divergence 3000 years ago.”

    Since Chinese and Japanese are unrelated languages, and 水 is just a character, not a word, borrowed from Chinese (a long time ago, but not 3000 years), this sentence is wrong in multiple ways. I think the author doesn’t really acknowledge the amount of borrowings that exists among langauges, especially for scientific terms.

Leave a Reply