Thread:Substantive changes (1)
From WikiPathways
The highlighted comment was created in this revision.
Miranda, thank you for the substantive changes to one of the most classic of pathways. Let us know when your editing is complete and we will review before updating its feature at wikipedia
That would be great if you could to that. There will be a supporting article explaining why we made these changes. We will submit this article this week. I do have some questions. The text under 'This pathway contains genes, proteins or metabolites with missing database annotations' is that automatically generated? I think I have most annotation except for the references to other pathways. Another question I have is that I now use "geneProduct", I am not sure if I should have used "protein"? Furthermore, the connection between reaction and geneproduct is "catalysis", but still the system seems to prefer gene IDs instead of protein IDs. And I cannot find a way to say which gene encodes for the protein. Last question, we looked up in ChEBI for most metabolites the version of the metabolite with the correct charge for the mitochondrion. We also have the HMDB IDs which I now used, but the automatic retrieval system then gives back the "neutral" versions of the metabolites in ChEBI. I do not seem to be able to annotate a metabolite with 2 IDs.--Mdstobbe 07:46, 3 April 2012 (UTC)
1. Yes it's automatically generated. You can delete the curation tag at any time, if you feel you've addressed as many of the issues as possible. Note that our bots apply new tags after every edit, so if you're planning more edits, wait until you're done to delete the tag. 2. Up to you whether you use "geneProduct" or "protein". Both are accurate. 3. "catalysis" is appropriate. The system does not have a preference, per se, it has an alias database with gene and protein IDs. You can choose whichever you find most effective. Downstream analysis tools should offer ways to convert/unify if necessary. 4. You can not annotate a datanode with 2 IDs. You can add the terms you've looked up as a comment on the node, but only 1 ID is accepted for mapping. And I recommend using the one in the database so we can support linkouts and ID mapping, etc.
Thank you for the answers. I tried now to annotate the geneproducts with the appropriate UniProt ID, but then in the "live" version it says that "Unable to load external references.", despite that this does work when I am the "editing mode". If I use the Entrez Gene IDs to annotate a protein then the external references can be loaded. A similar thing happens for metabolites, where HMDB IDs seem to work best. The curation tag "This pathway contains genes, proteins or metabolites with missing database annotations'" now only "complains" about the references to the other pathways andnd about the metabolite H+ and ubiquinol-10. I do not know why and how to correct this.
