One month ago today I put online a “General Overview of ChemSpider” presentation for everyone to review (of course it’s already well out of date). There were some interesting comments to that blog posting in terms of whether we would be making available a web API. As part of a collaboration we are presently developing a series of web services (more news about that soon). As part of this general direction and because of the recent interest in InChI (1,2) we decided to expose a series of web services built on a combination of OpenBabel and the InChI components.

A number of web pages exist exposing InChI-related capabilities including 1 and 2 . We are now exposing our own series of Web Services based on InChI as shown below.

InChI Web Services

On ChemSpider we have indexed over 17 million chemical compounds and have associated InChI strings and InChIKeys. The services include the ability to search for the Appropriate ChemSpider ID based on the InChI string and InChIKey. This is our first effort to offer programmatical access to ChemSpider and provide web services to the community. We might have missed some services with the InChI exposure here that you have interest in. Let us know. Also, since we are exposing other services right now if you have specific requests please let us know proactively!

Stumble it!

14 Responses to “Access to ChemSpider Web Services Starts – Initial Exposure of InChI Related Services”

  1. ChemSpider Blog » Blog Archive » Hacking PubChem - Technology Easy, Quality Difficult says:

    [...] Access to ChemSpider Web Services Starts – Initial Exposure of InChI Related Services 14 09 [...]

  2. Joerg Kurt Wegner says:

    Can you please provide all cross-services, means also something like
    SMILES2CSID
    SMILES2InChIKey
    etc.

    Maybe the page can be reorganised in a table format for seeing gaps in the conversion directions?

    Furthermore it would be great to have for all the From2InChI, From2InChIKey, InChI, and InChIKey an additional option in the service, which is
    search=[google,chemspider,emolecules,pubchem]

    Easiest is just to forward the searches to them, but it would be even greater to aggregate them together combined with ChemSpider searches? In other words, is it technical possible to get a kind of feed-aggregation/meta-search for chemical searches?
    This could be a very useful service anyway to prepopulate data entries for the structure upload process. This will avoid that people have to retype information from the other services, if you have not added that already, anyway?

    Why am I asking for this?
    1. If we do not got a special search site at Wikipedia, this might be a solution for creating a superservice for the non-uniform solutions out there. In other words, can you create a chemical Yahoo-pipe service?
    2. This would allow me to change my search boxes on my blog and on iGoogle with a centralized ChemSpider solution. The point is that additional information can be optional, by adding additional parameters to the request.

  3. Antony Williams says:

    Joerg, Sure, a full set of conversion services can be provided… but we’re afraid that they’ll grow as N^2. May be it’s better to simply provide a convert() operation taking formatFrom and formatTo arguments. Then we can examine the statistics of usage and can implement shortcuts for the most popular operations. What do you think? If they don’t get used at all then we haven’t consumed important cycles from an already overwhelmed ChemSpider team.

    Also we’re going to provide separate service for capabilities which can be extracted from OpenBabel (the list of those to be exposed is open still). This is hopefully not going to cause any legal issues of course based on our understanding of OpenBabel licensing.

    Please elaborate on your suggestion to search=[google,chemspider,emolecules,pubchem]. In particular – what format do you want to get results in? We can format our ChemSpider results (obviously), but what to do with those from other sites? Also, will there be any legal issues? At least one of the sites you list isn’t responding to requests.

  4. Developer at ChemSpider says:

    We’ve updated the list of InChI services and moved OBConvert related stuff to http://www.chemspider.com/OpenBabel.asmx

    Please DO NOT forget: these are Web Services, not web forms! Of course they can be used as web forms for testing purposes, but it isn’t their true destiny. We gonna provide web pages for all web services separately. Watch this space.

  5. ChemSpider Blog » Blog Archive » ChemSpider Team Exposes More Services - These Based on Open Babel says:

    [...] Community we stood on the shoulders of giants yesterday (the InChI team) and rolled our the InChI web services. Today we stood on the shoulders of a much bigger team of giants…all of those that have [...]

  6. Joerg Kurt Wegner says:

    1. With respect to the format I would guess that there should be an additional option like ‘format=xml,html,csv,smiles’. I would guess that other people prefer another output.

    2. the ‘from’ and ‘to’ parameter is a very good idea to reduce complexity, then you just need ‘from’, ‘to’, ‘format’ and ‘searchin=google+xyz+abc’

    3. web-service versus web-forms, well and? I act as simple user, which does not like to fill in forms, if I can avoid it. As said before, a simple workaroud might be to provide a save/load funtion for web-forms, where people could save/load already entered data. If you create by accident the same format by using a web-service, then you have already the connection. There are maybe more elegant ways doing this … any ideas from the community about this?

    4. Non-responding services and persons happen. Well, if they want to play alone in their sandbox, let them. I think I have a pretty good idea which one this might be. Anyway, the legal issue is a very serious one. And to be honest is not a golden rule to ‘break down barriers between services’ for making some progress. In other words, this is something that has to be worked out. If you can not get indexes from other services you can at least link to them. Though I would prefer an aggregation which could lead to an uniform output. Copyright and intellectual property statements could be added to the full search or to single entries of the search. In fact, this is even recommended!!! Honor to whom honor is due. Appreciate other services, if they do not the same to you. If they are not cooperating or responding, well, feel free to blog about it. The community is able to see who has done what and who has contributed what. And we all have to accept that services and people might have different goals! I personally see no point in not collaborating or at least talking about the potential intersections! If its about money, copyright, or appreciation there are tons of models to work with, not even talking about them is definitely not a solution.

    5. Another thing is that services do not like download robots and they might even block you and you can do the same. This is definitely the complete opposite of the recent goal, but this might happen … and again … I know that you are a talking, connecting, and hard working team, so I would be really surprised, if not at least some services respond positively to your request?

  7. Rich Apodaca says:

    I’ve written and documented a small Ruby library that uses the API:

    http://depth-first.com/articles/2007/09/17/hacking-chemspider-query-by-smiles-and-inchi-with-ruby

  8. Developer at ChemSpider says:

    I’ve added one more service for the sake of completeness: http://www.chemspider.com/Search.asmx

    It basically does the same as simple search at http://www.chemspider.com/Search.aspx

  9. ChemSpider Blog » Blog Archive » Structure Images on ChemSpider Tagged for InChI Searching. Web Service Enabled. says:

    [...] you’ve been frequenting our blog(s) you will have seen our passion for InChI adoption (1,2). A few months ago I started a discussion with Martin Walker (Walkerma), a very active member of [...]

  10. ChemSpider Blog » Blog Archive » ChemSpider Makes it to Second Life and to You Tube Courtesy of Hiro Sheridan says:

    [...] rezzer that accepts InChI, SMILES, and InChIKey identifiers. It takes user input and queries the Chemspider web services (for InChiKeys) plus two more of Rajarshi Guha’s web services (format conversion and [...]

  11. Jan says:

    Hi!
    I’m interested in converting InChIs in to 2D structures (mol files). Do you see any chance to include the 2D information in the short term? So far, only 0D Structures are replied.

    However, I appreciate the schemspider web servides very much.

    Regrds
    Jan

  12. Antony Williams says:

    Jan, Please use the online services page to convert InChI to structure. Visit http://www.chemspider.com/Services.aspx? and paste in your InChI and then Convert. You can then search the structure or download the structure

  13. Jan says:

    Hallo Antony!

    I knew this possibility already but was looking for web service offering that functionality of converting a InChI into structure. Chemspider renders the molecule behind the InChI in a manual kind of interaction, but in the web service that functionality is missing. Is there any particular reason?

    I know there are combinations of tools (ruby, CACTVS) around, but they all are not as handy and light weight as a web services would be.

    So, do you think, it could be possible to provide 2d structures in the future?

    I appreciate your comments very much!

    With kind regards
    Jan

  14. Antony Williams says:

    Yes…we can provide this service in the fairly near future. We need to finish some specific project aspects right now before we can move to this but yes it’s feasible.

Leave a Reply