Check current GO rules against previous implementation

[kltm]

Looking at the "Mike" script, there are some rules that may be useful that are not currently implemented.

http://geneontology.org/page/annotation-quality-control http://geneontology.org/page/annotation-quality-control-checks

Go through the latter and see what can be salvaged. This ticket is to be closed once the above has been reviewed.

[pgaudet]

http://geneontology.org/page/annotation-quality-control-checks

Implemented

• GO_AR:0000001 Basic GAF checks: see #715
• GO_AR:0000002 No 'NOT' annotations to 'protein binding ; GO:0005515' -- GORULE:0000002: not implemented ?
• GO_AR:0000003 Annotations to 'binding ; GO:0005488' and 'protein binding ; GO:0005515' should be made with IPI and an interactor in the 'with' field -- GORULE:0000002: implemented as report
• GO_AR:0000006 IEP usage is restricted to terms from the Biological Process ontology
• GO_AR:0000019 Generic Reasoner Validation Check

Approved

• GO_AR:0000004 Reciprocal annotations for 'protein binding ; GO:0005515'
• GO_AR:0000005 No ISS or ISS-related annotations to 'protein binding ; GO:0005515'
• GO_AR:0000016 IC annotations require a With/From GO ID
• GO_AR:0000017 IDA annotations must not have a With/From entry
• GO_AR:0000018 IPI annotations require a With/From entry
• GO_AR:0000020 Automatic repair of annotations to merged or obsoleted terms
• GO_AR:0000007 IPI should not be used with catalytic activity molecular function terms
• GO_AR:0000008 No annotations should be made to uninformative high level terms

Proposed

• GO_AR:0000009 Annotation Intersection Alerts
• GO_AR:0000010 PubMed reference formatting must be correct
• GO_AR:0000011 ND annotations to root nodes only
• GO_AR:0000013 Taxon-appropriate annotation check
• GO_AR:0000014 Valid GO term ID
• GO_AR:0000015 Dual species taxon check

[pgaudet]

@vanaukenk @ukemi We should review these on Wed to make sure they are still what we want.

[ukemi]

I suggest we triage: ones we certainly want to keep, ones we want to certainly discard and ones we want to reconsider based on GO-CAM modeling.

[vanaukenk]

Yes, let's discuss on Wednesday. I think we should keep GO-CAM models in mind for reviewing most all of them, actually, and if needed, create a model to check and/or illustrate the implementation.

[dougli1sqrd]

Has this been appropriately reviewed?

So says @kltm

Go through the latter and see what can be salvaged. This ticket is to be closed once the above has been reviewed.

[dougli1sqrd]

Maybe we should merge this with #670; @pgaudet can you make sure there is no missing information?

[pgaudet]

We have everything in #670.

Thanks!