Difference between revisions of "WormBase-Caltech Weekly Calls"
From WormBaseWiki
Jump to navigationJump to searchLine 96: | Line 96: | ||
*Phenotype-2-GO mapping: What is the purpose? How does it work? | *Phenotype-2-GO mapping: What is the purpose? How does it work? | ||
**Broad vs specific term mapping; how good is the mapping? | **Broad vs specific term mapping; how good is the mapping? | ||
+ | |||
+ | |||
+ | CSHace merge with CITace? | ||
+ | *PCR_product info discrepancies | ||
+ | *CITace data was overwriting CSHace data, but changed | ||
+ | *8 classes of data in CSHace | ||
+ | *CSHace data had never been read into CITace minus | ||
+ | *Proposal: read CSHace data into ACEDB and then read in CITace minus data (which would overwrite CSHace data with CITace data where there is a conflict) | ||
+ | *Can we assume that all CITace data is better/newer/more-up-to-date than CSHace data? |
Revision as of 19:04, 11 October 2012
2012 Meetings
October 4, 2012
FlyBase SAB topics
- Genome Space - can integrate data in different formats
- Cloud-based data integration
- File conversion done automatically - no need to write scripts
- FlyBase talking about joining Genome Space?
Protein-to-GO Tool
- Ranjana and Kimberly performing data checks for testing
- If file can get to Rachel and Tony, maybe can discuss this weekend (at GO meeting)
GO Annotator Tool
- Kimberly looking over
- James will make a couple slides to introduce tool and provide demo
- 1-minute demo: can have file in XML/HTML format, can highlight, annotate a sentence, and save annotation as a link
- Demo: simpler = better
- Prepare for live feedback and discussion
- 1-minute demo: can have file in XML/HTML format, can highlight, annotate a sentence, and save annotation as a link
Curation Status/Statistics Tool
- How many papers have given data types?
- How many papers have been curated, how many not?
- How many objects/connections do we have of a given type?
- How many objects per paper (average, distribution?)?
- Estimated number of objects/papers exist vs. how many we have curated?
- Do we care about types of flagging? Curator first pass, author first pass, etc? Yes
- All flagging types should be shown; combined and individual statistics would be useful
- Interactive form vs. static page showing curation/flagging statisitcs
- Tracking negatives (true vs false)
- Data types in OA vs not (microarray, Protein-to-GO output)
- Microarray data - Wen can write script to generate stats for microarray curation
- Curation stats per paper (Raymond): write comments in a remark field; traceable, transparent, available
Process pages
- GPML - may be able to automatically map Postgres data to GPML
- WikiPathways has color, formatting, labels, etc. so we can define types/views of different relationships (should follow standards)
- AWC-ON/AWC-OFF sample page: only single connection type; can be many different types, as we define it
- "Too many arrows" editorial?
Physical interaction curation
- On interaction OA (Tab2) physical interactions; 'Colocalizes', etc.
- Need to establish our data exchange with BioGRID; get BioGRID data in Postgres/OA (daily cronjob?)
- Interaction model was modeled to be compatible with BioGRID's data
- Revive curator first pass for physical interactions? To tag papers (in the meantime)
October 11, 2012
Molecules
- Changing to WBMolIDs
- OA dumpers need to be modified
GO Meeting
- How to make GO more expressive/inclusive
- How can GO represent more of the biology
- GO Extensions: LEGO
- Deeper annotations can help develop the "big picture"
- Phenotype-2-GO mapping: What is the purpose? How does it work?
- Broad vs specific term mapping; how good is the mapping?
CSHace merge with CITace?
- PCR_product info discrepancies
- CITace data was overwriting CSHace data, but changed
- 8 classes of data in CSHace
- CSHace data had never been read into CITace minus
- Proposal: read CSHace data into ACEDB and then read in CITace minus data (which would overwrite CSHace data with CITace data where there is a conflict)
- Can we assume that all CITace data is better/newer/more-up-to-date than CSHace data?