1 Jun 2021 PAINT Conference Call: Difference between revisions

Revision as of 18:58, 26 May 2021

Need to define specs; proposal (Pascale) PAINT would load GP2term relations so PAINT curators would be able to select more precise annotations for propagation.

Recording any drastic changes like increases or decreases
Where would the release notes be available?
- Example URL: http://data.pantherdb.org/ftp/downloads/paint/current/README (doesn't exist yet)
- Or just roll into GO release notes? Ex: https://github.com/geneontology/go-site/tree/master/releases/2021-05-01

How they currently affect PAINT objects
- Experimental annotations - considered valid to use for IBD evidence if no TC violation
- IBDs - will be obsoleted if IBD itself violates a TC or if it loses all experimental evidence due to TC checks in step #1
- TCVs - created in the family trees from the TCs and used to block propagation of IBAs
Tool logic vs validator (batch process) logic?
- Should rules and effects be consistent between curator actions in the tool vs the monthly validator runs?
  - Ex: Curator is allowed to create an IBD in violation (with a warning) but validator doesn't allow this and will obsolete later. Maybe this is a nonsensical example?

@@ Line 9: / Line 9: @@
 ===NOT annotation issues===
 https://github.com/geneontology/go-annotation/issues/3701
+===IBA Release notes===
+* Recording any drastic changes like increases or decreases
+* Where would the release notes be available?
+** Example URL: http://data.pantherdb.org/ftp/downloads/paint/current/README (doesn't exist yet)
+** Or just roll into GO release notes? Ex: https://github.com/geneontology/go-site/tree/master/releases/2021-05-01
+===Taxon constraints in PAINT===
+* How they currently affect PAINT objects
+** <b>Experimental annotations</b> - considered valid to use for IBD evidence if no TC violation
+** <b>IBDs</b> - will be obsoleted if IBD itself violates a TC or if it loses all experimental evidence due to TC checks in step #1
+** <b>TCVs</b> - created in the family trees from the TCs and used to block propagation of IBAs
+* Tool logic vs validator (batch process) logic?
+** Should rules and effects be consistent between curator actions in the tool vs the monthly validator runs?
+*** Ex: Curator is allowed to create an IBD in violation (with a warning) but validator doesn't allow this and will obsolete later. Maybe this is a nonsensical example?
 [[Category:PAINT]]