EC xref cleanup #12988
Comments
|
I did this last year: 4 got away, because they were incorrectly formatted: Two ended in period, one was missing a period, and the other has two lines glommed into one. This could have happened via hand editing (most likely), or perhaps by pasting windows characters into the xref box in OE. In the intervening time, people have been adding dash xrefs back in: I will remove these, and add a check to stop this |
|
In r38555 I removed all dashes, and I added a EC dash checks to the script. They prohibit dashes in EC primary xrefs. They are allowed in def xrefs (though this is dubious). |
|
I created a repo for doing a deeper dive on EC and other MF xrefs: https://github.com/geneontology/go-xref-analysis |
|
I created a report of all ECs mapped to sub/superclass pairs: https://github.com/geneontology/go-xref-analysis/blob/master/redundant-xrefs-ec.tsv Someone give me the word and I can removal all the ones mapped to a child |
|
I looked at this a while ago. |
|
I can start to go through with removing the grouping ecs from the more granular terms, but some of the 4 digit ones need to stay. These really can't be used as decent def refs either; I would prefer to at least replace them with GOC:curator or something. |
|
There are 183 EC's in the list Chris created. I have removed seveal today, so it's not a BIG list to do manually; I'm specifically going for the 3 digit ecs; I'm not touching anything with 4 at the moment. I think they can be used for several, like EC 2.6.1.42; |
|
Removed a bunch today; will continue |
|
More today. If no replacement ec, need a PMID or will have to use GOC:hjd |
|
Hi @cmungall |
|
@balhoff Can you get the list of GO terms with partial ECs ? (no rush) |
|
Also note I am still removing partials from the plant enzymes and replacing with RHEAs (several are pending). |
|
@pgaudet do you still need a list of GO terms with partial ECs? |
|
Yes :) |
|
Note: i have been slowly cleaning these out from the plant enzyme project. Sometimes the metacyc xref has a full EC to use. When that happens, I can then try to get a RHEA to use as a def ref. Also, I can sometimes replace the def ref if it is an EC with one or more of the pubmed ids supplied in the Metacyc xref especially if it is used for enzyme purification. You can't just blanket remove these because it sometimes leaves NO def ref. |
|
I work using the original #12988 ticket that Chris originally supplied a spreadsheet for me. |
No description provided.
The text was updated successfully, but these errors were encountered: