20 January 2015 PAINT Conference Call

January 20th, 2015 PAINT call minutes

Participants: Suzi, Paul T, Moni, Huaiyu, Rama, Karen

Agenda

Huaiyu updated the PANTHER v10 status.
- Trees are built. In the process of loading to the database and publish. Probably will take another month or 2 to release.
- 104 genomes in this release. Family number increases to around 12,000.
Suzi suggested to allow curators to display branches of tree, eg., only the plant sequences, by using a webservice with the prune tree algorithm.

Paul visited Suzi's group last week. There were discussions about making two changes to PAINT to improve the curation process.

Allowing partial loss of function at an ancestor node. When a specific function is lost, a more general function may still exist. Therefore, when a specific GO term is annotated as NOT, a more general GO term is still allowed to be propagated.
Constraint check. Curators are allowed to make the annotation that may violate the constraint rules, and build the model for the entire family. For example, the curators may ignore the taxon constraint, or make a positive annotation when there is a negative annotation from literature curation. When the annotations are saved, e.g., when the Save button is clicked, the PAINT will make a final constraint check to find potential conflicts, and display the waring. The curator must address each of them by either going back and correct the annotation or recording reasons in the notes to allow the annotation to be saved.

Suzi will implement these after she finishes other PAINT features, such as forward track annotation.

This is about the email Rama sent out on Jan. 15th, 2015. It is PTHR10130. There are NOT annotations in the gaf file on leaf sequences with IRD, but not in the ancestor node.
Paul T looked at the tree and concluded that the NOT annotation on leaf is correct. There are good notes in the notes file. The original gaf file in snv was reviewed and found that it had the same gaf file. There might be a software problem recording the NOT annotation back then. Suggested Rama to reannotate the tree to correct the problem.