Several BiGG SBML models are invalid and cannot be easily read

[smoretti]

Hi

Several BiGG models are invalid in SBML format:

• "listOfFluxObjectives cannot be empty" e.g. iMM1415, iSynCJ816, iAT_PLT_636, iRC1080 Some models have no objective functions (see #333), but other like iMM1415 had one before.

• rdf:resource URI containing unsafe characters, complexifying interpretation (see https://www.w3.org/Addressing/URL/4_URI_Recommentations.html) e.g. iJN678, Recon3D, iAB_RBC_283, iCN900 Some models contain spaces in the sbml:annotation rdf:resource URI:

  • rdf:resource="http://identifiers.org/refseq_synonym/trxR in PMID:23222730"
  • rdf:resource="http://identifiers.org/refseq_synonym/MCT 3" Is it different from rdf:resource="http://identifiers.org/refseq_synonym/MCT3" in the same entry?
  • rdf:resource="http://identifiers.org/refseq_name/braG or livF" This one mean rdf:resource="http://identifiers.org/refseq_name/braG" or rdf:resource="http://identifiers.org/refseq_name/livF" ; or something else ?

Thanks

[mephenor]

This most likely stems from ModelPolisher currently not being used to polish the models, due to issues with the latest release, as both the listOfFluxObjectives and refseq_ issue should be handled by it. I am currently working on some remaining issues with regard to annotation and performance and adding thorough testing to at least the core parts so that it becomes usable once again.

[smoretti]

Thanks Waiting for your fixes

[smoretti]

Any news for this fix?

[mephenor]

Currently fixing small issues still existing with ModelPolisher and running polished Models through the Validator to check for further issues. If everything goes well, we could have a ModelPolisher release tomorrow and then the BiGG Models team can decide, if they want to use it once again in its current state.

The refseq_* annotations listed cannot be resolved, so they will be removed, if you run the model through ModelPolisher.