14 DEC 2010 RefGen Phone Conference (Archived)
- Mike L
- Mary Dolan
- Curators: spend time between now and the end of January curating families. We hope each curator can spend around 40-80 hours over the next 6 weeks - this figure is just to give an objective and may not be achievable. In any case, curators should keep track of how long it takes them for the annotation of each family in the Tree_annotation_progress page, so that at the end of the exercise we are able to estimate the resources needed for this project.
- Suzi: update the PAINT_User_Guide to describe how to change the mirrors used by PAINT.
- Pascale: pick a family to be annotated by everyone and discussed in January.
- Chris: check if there's a way to make the text from the evidence.txt file get copied in the wiki automatically.
"Fast pass" curation effort
We are beginning to PAINT the 418 remaining families that have been RefGenome targets. For this project, the goal is to use PAINT to curate quickly, capturing the most important aspects of the biology of each family. Broadly speaking, we hope to spend on the order of one hour per family on average. Obviously this can vary widely. The general rule is that if it will take too much time to determine whether or how to make an annotation, then do not make it. We want to spend on the order of an hour per family, not several days.
An initial set of families has been divvied up: http://gocwiki.geneontology.org/index.php/Tree_annotation_progress
Linkouts are provided to pages for individual families. Most of these pages are new. Please record annotation questions in them using established structure. Example of structure: http://gocwiki.geneontology.org/index.php/PTHR12027
There is no log in now, and files will be committed using a common login. Curators should write their names in the Evidence file (following the format that will will specify in the Paint User Guide).
PAINT software update
- Most recent release : beta33
- Fixes a bug that when there are annotations to both a specific and a more general term, the proteins with the most specific annotation were not inheriting the more general annotation. However, if one deleted the more specific annotation, then the annotation to the most general term was not restored. This is now fixed.
- IN PAINT beta33 the dates annotations were made are now visible.
- ISSUE: PAINT beta33 has a problem on PCs with the latest update of Java (Java 6 revision 22) in that it doesn't fetch the annotations from the GO database. This is hard to resolve because the problem is not in PAINT. Looking for solutions.
- ISSUE: There is an issue with the FIND function: if a term is found in another ontology, it doesn't 'jump' to that other view, but appears as if there is no result.
- There is a PAINT user guide on the wiki: http://wiki.geneontology.org/index.php/PAINT_User_Guide
- Curators should connect to the closest GO mirror for faster downloading of data. PAINT user guide should explain how to connect to the closest GO mirror.