I’ve blogged previously about us adding safety and toxicity data to ChemSpider. We are busily sourcing new information from other data sources to add information and in the past couple of days we have added NIOSH data as it is a rich source of additional safety information. For example, the record for 1,2,3-trichloropropane shows:

  • First Aid: Eye: Irrigate immediately Skin: Soap wash Breathing: Respiratory support Swallow: Medical attention immediately

  • Exposure Routes: inhalation, skin absorption, ingestion, skin and/or eye contact

  • Symptoms: Irritation eyes, nose, throat; central nervous system depression; in animals: liver, kidney injury; [potential occupational carcinogen]

  • Target Organs: Eyes, skin, respiratory system, central nervous system, liver, kidneys Cancer Site [in animals: forestomach, liver & mammary gland cancer]

  • Incompatibilities and Reactivities: Chemically-active metals, strong caustics & oxidizers

  • Personal protection and Sanitation: Skin: Prevent skin contact Eyes: Prevent eye contact Wash skin: When contaminated Remove: When wet or contaminated Change: No recommendation Provide: Eyewash, Quick drench

Some additional examples are here: Temefos, Warfarin and Allyl Alcohol. Note that each of these also has a coincident extract from Wikipedia. We are therefore integrating Wikipedia articles, safety, toxicity, experimental and predicted properties. Our plan for semanticising and integrating the chemistry web is clearly well underway.

Stumble it!

9 Responses to “Adding More Safety and Toxicity Data to ChemSpider”

  1. Cameron Neylon says:

    This is great news and will hopefully feed into some things we are planning to do with our lab blog system in our planned roll out to our new lab. Linking back to safety data will be very helpful. Is there any risk here that you are setting yourself up as an authority though? And therefore the potential for being held responsible for the use of the information?

  2. Antony Williams says:

    I’d welcome any input from you regarding what more we can do on our side to enable your lab blog. You’re likely aware of the work we have done to support JC’s Open Notebook Science project?

    See: http://www.chemspider.com/blog/usefulchem-open-notebook-science-now-has-a-dedicated-chemspider-site.html

    If we can do similarly for you to assist in your project please let me know.

    One of the errors with any of the aggregating systems, whether its PubChem, eMolecules, ChemSpider or Wikipedia, is that calling them authorities is not appropriate. People DO though. Of the four listed I would trust Wikipedia far ahead of the others. I will blog tonight about how at least one type of error can occur and give rise to misassociation of data. We will be adding an appropriate disclaimer label to the Supplementary information shortly. The Disclaimer has been online since we went live: http://www.chemspider.com/Disclaimer.aspx

  3. Cameron Neylon says:

    Antony, we will definitely keep you in the loop going forward. We’re not yet at the stage of having anything as coherent as JC’s material but hopefully we are heading in that direction. One thing that could be interesting is having depositors as providers of safety assessments for specific compounds (although again there would need to be appropriate disclaimers). Would be very interesting to see whether it would be possible to build a community aroud the safety information to comment and encourage best practise. As well as dealing with the issues of differing regulations in different places.

  4. Antony Williams says:

    ChemSpider is ALL about Building a Community for Chemists and I would be honored to work with you to encourage best practices around safety information. Ready when you are…

  5. Physchim62 says:

    I am more than wary about the sort of safety information you are adding. For a start, let me say that we have been systematically removing this sort of data from Wikipedia for about the last three years.

    Take the example of chloroethyl ethyl sulfide, CASRN 693-07-2. This is a commercially-available compound (which I have used personally), but not included in any of the regulatory databases. As such, it wouldn’t have any safety information of the type “wear face protection”, “handle with care” etc. It is also a potent vesicant (blistering agent), as a comparison of its structure with sulfur mustard might suggest. The lack of safety information doesn’t make this compound any less dangerous, but the inclusion of “generic” safety information for other compounds might suggest this.

    On Wikipedia, we have been careful to include the safety information in our possession (within the limits of a collaborative project, obviously) without implying that we will always have such information for a given compound. It is not an easy job, but I think it is worthwhile in the long term. Otherwise, you will fall into one of the traps: either “every chemical is dangerous” or “we’re not going to risk saying anything”.

  6. Antony Williams says:

    Physchim62 – what’s the wariness arout? Is it because the data are inaccurate? Are of no value? I’d love to get pointed to the Wikipedia page discussion about the decision regarding why the data are being removed from Wikipedia. Overall the number of people who have responded to this work (small number) are all positive but I respect your opinion and am definitely up for a discussion anytime!

    I looked for CASRN693-07-2. It’s here: http://www.chemspider.com/Chemical-Structure.12210.html on ChemSpider but couldn’t find it on Wikipedia. Where is it on Wikipedia and then we can link to it.

    What I hear you saying then is it is better to not have ANY safety information for any compounds at all than to have it for some? Wouldn’t this be the same for Wikipedia where every compound must have safety information. Maybe I am missing the point?

    I don’t think users of ChemSpider can expect us to have safety information for every compound on the system. In any case there is a disclaimer on Supplementary Information. I agree with you re. the traps…everything is dangerous and nothing is dangerous. We are doing what we can. If you have a suggestion about how we should improve the effort I am all ears and would love to gain from your experience and not rework.

  7. Cameron Neylon says:

    I’m with Antony on this one I think. We desperately need to get away from the extreme kinds of safety data we have (everything is dangerous, no useful information is available in case of legal action) and provide useful information that actually helps people to make an informed decision.

    What I was thinking of was depositing the actual procedures (risk assessment and COSHH forms in UK language) that we use when handling specific compounds. With appropriate disclaimers that we are not authorities and that no-one should rely on the information. Some of it will be wrong or incomplete, but if people tell us that, and if we tell other people, then everyone is safer.

    I agree it is important to make it clear that absence of evidence is not evidence of absence (of a risk) but surely this can be made clear;

    ‘We do not have any safety data on this compound. It cannot be assumed to be safe. You should always make your own assessment of the risks posed by any compound that is appropriate to your local regulations and policies’.

  8. Antony Williams says:

    Cameron, I looked up COSHH online. I was especially impressed to see this http://www.hse.gov.uk/welsh/index.htm especially since I once upon a time had decided to be a Welsh teacher as a career (growing up in a small village in North Wales!)

    I couldn’t find details of the nature of NIOSH. CAn you point me to examples of COSHH forms for me to review?

    BTW, what a great statement basis you have provided here “‘We do not have any safety data on this compound. It cannot be assumed to be safe. You should always make your own assessment of the risks posed by any compound that is appropriate to your local regulations and policies’.” Maybe it’s appropriate to put this as Supplementary Info for EVERY compound OTHER than those where Safety information already exists. I’m thinking…

  9. Cameron Neylon says:

    Antony, I can email you some examples but some of the ones we already have up on our lab book can be found here:


    You will also see our disclaimer at the bottom :)

    One thing that has been interesting comparing our practice to Jean-Claude’s in the Open Notebook area is the area of safety. The notebook above had the safety data put in after a visiting postdoc point out ‘It’s a lab book, it should have the risk assessments’ and this is indeed standard practice in chemistry in the UK (perhaps I should say ‘best practice’).

    The format of COSHH and risk assessments vary from place to place but the principles are the same. My lab manager has just been on a NIOSH course so will ask her for a pointer to the latest online info (probably toll access unfortunately). General COSHH info is at http://www.hse.gov.uk/coshh/index.htm

    I would definitely put some sort of disclaimer as the minimal safety data for any compound. Good chemical practice after all is to treat any new unknown compound as thought it could be toxic.

Leave a Reply