QCQA next call: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
mNo edit summary
 
(11 intermediate revisions by 2 users not shown)
Line 1: Line 1:
=Preparing for GOC meeting =


=Annotation redundancy - non-experimental annotations=
Action points from GOC meeting
**Can we come up with a proposal for the Montreal meeting?
* Using Matrix more widely for QC
**http://wiki.geneontology.org/index.php/2018_Montreal_GOC_Meeting_Agenda#Handling_redundant_information
* See whether FlyBase Annotation Saturation strategy can be used at a larger scale


==Discussion: Annotation redundancy - non-experimental annotations==
* Remove all PINC annotations?
* We need to define what constitutes a redundant information, WRT to sources, evidences and references
* There is a ticket that explains the strategy that will be taken for this: see https://github.com/geneontology/amigo/issues/43 and https://github.com/geneontology/amigo/issues/440
* '''AI: Rules for flagging redundant annotations need to be documented - Chris'''
* https://github.com/geneontology/go-annotation/issues/2060
* https://github.com/geneontology/go-annotation/issues/1879


Different types of redundancy
=UniProt annotation issues=
* identical experimental annotations from different sources https://github.com/geneontology/go-annotation/issues/1404
* Figure out a way to export users from UniProt annotations
* inferred and IEA, ISO, ISM, IBA  annotations  when EXP exist
* Could we have the ORCID for everyone ?
* Redundant  inferred annotations from GOC F-P links pipeline
    e.g.  https://github.com/geneontology/go-site/issues/576
* TAS /NAS (special case)
 
There is no use case for users of MODs and AMiGO to see automated annotations when EXP annotations are present.
If somebody wants a complete set of annotations from a mapping resource these could be made available separately.
 
==New error reports - status==
http://snapshot.geneontology.org/reports/gorule-report.html
https://github.com/geneontology/go-site/issues/709
 
=How to blacklist papers=
Pascale
* List/form template of excluded papers or combinations of PMID/Gene/GO
 
= Update: Unmaintained annotation sets=
https://github.com/geneontology/go-annotation/issues/1724#issuecomment-390136730
* '''AI: check status with Tony - Unmaintained annotation sets'''
 
TIGR, JCVI, PAMGO ISS annotations: Michelle:
The ISS annotations from at least TIGR (and should be also JCVI and PAMGO) were all manual. They were either matches to HMMs or based on pairwise alignments. We used IEA for any annotations that were automatic from our pipeline and not reviewed. I assume JCVI continued that process after I left - at least until they stopped using their pipeline and shifted completely to using the NCBI PGAP tool (prokaryotic genome annotation pipeline).
I think we (and hopefully PAMGO) were pretty good about using the GO_REFs to indicate whether it was HMM or pairwise. The problem with replacing our HMM annotations with InterPro2GO mappings is that we made much more specific annotations based on HMMs than what the InterPro mappings often do. I'd hate to lose those but I understand your desire to keep the annotations current.
 
= Reported from the previous meeting =
* GO-CAM QC plan
** including SynGO, > 90 % of the production models
 
 
=To revisit later=
==SPKW hierarchy - at odds with transitivity as found in GO ontology?==
**https://github.com/geneontology/go-annotation/issues/2036
**emailed Chris, Seth about help with querying the GO database for non-redundant SPKW-based annotations (2018-08-16)
** Kimberly: WB looking into the information added by SPKW in WormBase. From that evaluation we can perhaps decide whether we want to keep them. 






[[Category:Quality Control]]
[[Category:Quality Control]]

Latest revision as of 07:06, 28 June 2019

Action points from GOC meeting

  • Using Matrix more widely for QC
  • See whether FlyBase Annotation Saturation strategy can be used at a larger scale
  • Remove all PINC annotations?

UniProt annotation issues

  • Figure out a way to export users from UniProt annotations
  • Could we have the ORCID for everyone ?