#2 Incoherency in fma_obo.obo file

closed
nobody
None
5
2009-06-17
2008-03-05
Robert Baud
No

I just download this week (3 March 2008) the file fma_obo.obo from http://www.obofoundry.org/cgi-bin/detail.cgi?id=fma_lite

Working on this file I found the following 8 entities, which occur twice in the hierarchy. In fact they are each time different entities with the same FMAid, which is supposed to be a unique identifier.

The problematic 8 FMAid are the following:

13065, 33455, 36611, 58401, 65162, 79222, 227947 and 239796

Can you tell me if I am right? If any corrected file in the future, let me know.

Thanks in advance and anyway congratulations for the FMA.

Discussion

  • Chris Mungall

    Chris Mungall - 2008-07-24

    Logged In: YES
    user_id=254723
    Originator: NO

    Hi Robert

    You are correct, and this is very problematic. I am not sure if this is a problem with the source FMA or with the conversion.

    In any case, the fma_obo is lagging behind the current FMA. I am working on a new conversion to obo and owl-dl just now (note: I didn't work on the original conversion)

     
  • Chris Mungall

    Chris Mungall - 2008-07-24

    Logged In: YES
    user_id=254723
    Originator: NO

    draft of an email I am about to send to obo-anatomy (which you should be on if you are interested in this):

    The fma_lite version which has been available here:
    http://www.obofoundry.org/cgi-bin/detail.cgi?id=fma_lite

    is lagging behind the full FMA, now on version 2. Version 2 includes various improvements, particularly the removal of erroneous part_of relations which only held in a SOME-SOME rather than ALL-SOME fashion.

    I have an *alpha* version of a translation to obo available here:

    http://obo.svn.sourceforge.net/viewvc/obo/fma-conversion/trunk/fma2_obo.obo?view=log

    The conversion is based on the Stanford conversion of the frames version to the FMA to OWL-Full:
    http://www.bioontology.org/wiki/index.php/FMAInOwl

    Note that the translation is different from the one used to create the current fma_obo.obo; in particular:

    - the fma2 translation includes a wider range of relations:

    systemic_part_of
    regional_part_of
    constituitional_part_of
    (all subrelations of part_of)

    I can add more relations to the translation on request

    - there are no artificial template metaclasses, in line with OBO principles

    There are still some outstanding issues, in particular:

    - A number of classes full under "general anatomical term". These should be relegated to the status of synonyms. As yet I am unsure how to do this from the OWL Full translation.

    I can make an OWL-DL version available on request (note that this will differ substantially from the OWL Full translation, which relies heavily on metaclasses)

     
  • Chris Mungall

    Chris Mungall - 2009-06-17

    fixed in fma2.obo (current release)

     
  • Chris Mungall

    Chris Mungall - 2009-06-17
    • status: open --> closed
     

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.





No, thanks