WormBase-Caltech Weekly Calls May 2014
From WormBaseWikiJump to navigationJump to search
- 1 May 1, 2014
- 2 May 8, 2014
- 3 May 15, 2014
- 4 May 22, 2014
- 5 May 29, 2014
May 1, 2014
- Curation status freezes have been collected (daily?)
- These can be used for report
- Review grant proposal: milestones, quarterly updates, etc.
LEGO - Paul Thomas' visit
- We want to work on/think about importing WB GO annotations into LEGO format
- Is .OWL format the final destination of LEGO data for us? Yes, probably
- Will we be able to automate the process and eventually port to WikiPathways or other pathway visualizer?
- Extended evidence codes? For capturing methods?
- We need some trial LEGO curation to start working out the process
- There is a communal desire to achieve a common format for pathways, whether it be SBML, SBGN, LEGO, etc. or some hybrid
- We would like to play around with how to import/interpret/export .OWL files for our data
- Topic curation: a single (or pair of) curator(s) focusing on pulling together and drawing a pathway for a topic
- Labs adopting a pathway? Outreach, community annotation
- Reaching out to experts to confirm or proof read our pathways
WormBase Ontology Browser (WOBr)
- Getting closer to completion
- Working out how to handle different relationships, e.g. "part_of", "has_part"
May 8, 2014
- We may be able to use BioTapestry (from Davidson group) to visualize gene regulatory interactions
- Xiaodong testing out
Li et al 2004 Interactome
- Mostly sorted out now
- Missing gene references have been filled in
- "Noncore" interactions will be omitted from website
- Almost perfect correspondence between the published dataset and WB dataset
- Still making progress
- Incorporating life stage and disease annotations
- There appeared to be missing gene-phenotype annotations; Raymond contacted Kevin Howe
- Raymond will try to work on from Taiwan
- Paul Davis deposited the most recent models revision (1.404) in the models CVS repository
- Curators should test their model changes
- ?Variation and ?Construct models still need work; data needs to be conformed to fit new models
- Questionable whether all the changes for ?Construct and ?Variation models and data can be made by the next upload (May 27th)
- ?Expr_pattern model being modified soon to accommodate annotation of specific conditions; working to be compatible with LEGO
Curation Progress Report
- Each curator can use the Curation Status Form to generate numbers for their datatype(s)
- Wen can install relevant versions of ACEDB (WS238 - WS243?) to pull annotation numbers
May 15, 2014
- Paul will send around format during or after the call
- Whole contribution from Caltech (bio-curation) should be about a page
- A few sentences on each curation topic
- Some mention of models, WOBr, topic curation
- Next Thursday (May 22nd) local deadline (official deadline June 1)
- WS238 - WS243, March 2013 (beginning curation) to March 2014 (end of curation) 12 months/1 year
- Curation Status Form has stats from Aug 1 2013, we can look at numbers from Aug 1 2013 to May 1 2014, 9 months, can be extrapolated to a year
- Xiaodong sent around some sample gene regulatory network diagrams based on WB data
- How do we want to use this?
- To what extent do we need time and place (spatio-temporal) information for a transcriptional regulation to use BioTapestry?
- Can we generate spatial-agnostic/temporal-agnostic diagrams?
- Xiaodong will look into curation numbers for annotations that include spatial info, temporal info, or both
GO Meeting at Caltech
- Need to reserve a room for next Tuesday/Wednesday, 10-15 people, all day (if possible)
- Tech-oriented discussions of LEGO, GO, Textpresso
May 22, 2014
- Orion/LEGO/Textpresso meeting
- Progress Report
- GO curation will eventually morph into LEGO curation
- Can we incoporate other WormBase data into LEGO? Hopefully, we'll have to see
- Would be good to have OA (e.g. Interaction OA) feed directly into LEGO/Orion/Noctua
- We are building pathways with various tools; long term curation goal not clear yet
- How could we represent other data in LEGO? Phenotype annotations? Expression patterns? Genetic interactions?
- Connecting tools like PAINT, OA, Textpresso, etc. to LEGO
- LEGO team worked with Juancarlos to get some LEGO tools/infrastucture (Noctua, Barista, Amigo) running at Caltech
- Mechanisms to seed LEGO diagrams is a priority
- LEGO team working with Michael to interface with Capella (Textpresso tool)
- Curators should send their blurbs to Chris (in any format)
- Chris will send plain text to Paul S.
- Human diseases connected to phenotypes via genetic variants in various model organisms
Athena is crashing
- Wen having trouble with Athena; difficult to reboot
- Data is backed up; Wen and Raymond will work on replacing Athena
- Citace upload to Wen by Tuesday morning
- Still waiting on curator timestamps and a dump of partially curated papers from BioGrid
- Will still need to compare genetic interaction datasets
- For large scale genetic interaction datasets in both databases, we will favor the WB version, import others (best mapping of terms possible)
2004 Vidal Interactome Yeast-Two Hybrid Data
- Chris has nearly finished sorting out official interactions from those removed
- Once done, we can import from later interactome datasets
May 29, 2014
- Progress Report
- Citace upload
- LEGO connected to OA/Postgres -- Oh, is _that_ what we're talking about ? I thought it was about writing an interface that looked like the OA that worked on their data, not on postgres at all -- J
- Can we link curation in the OA to LEGO model?
- Still needs to be worked out
- Paul putting together report
- We need to discuss priorities
- Wen received new phenotypes file
- Upload to Hinxton will proceed as usual tomorrow
- Wen will send around a new upload summary today
Cis-regulation interactions visualization
- Now that we have broken out cis-regulation events into binary interactions, we can tell Abby to display in Cytoscape on the website
- We will use Cytoscape now, maybe consider BioTapestry later
Scientific Advisory Board meeting
- Hinxton, UK, October 5-7, 2014
- Need to consider what types of numbers and coverage we want to cover
- e.g. Exppression patterns are caught up, behind on concise descriptions but we have a plan...
- We would like to acquire better coverage of allele-phenotype/RNAi-phenotype annotations
- Percent coverage, change in coverage percentage (are we keeping up? are we caught up?)
- LEGO curation approach: WB data modeling to capture LEGO type curation
- Biological Topics, WikiPathways, Reactome
- WormBase Ontology Browser
- Proportion of papers covered, # objects curated
- Text mining support to fill in missing information
(Semi-) Automated Concise Descriptions
- Primarily for genes with very little existing information (e.g. no papers)
- Proposed tool could pull up basic info on gene: orthology, expression clusters, RNAseq data
- Expression cluster info could provide info under what conditions a gene is up- or down-regulated
- Is degree of confidence a concern? Are there confidence weights that we should apply? Thresholds for making some statements?
WormBase Ontology Browser (WOBr)
- Currently we are still trying to adapt AMIGO 2 to WormBase data
- Amigo is Gene Ontology centric; to what extent does it work for us?
- How do we best merge Amigo GitHub code with WormBase code for further WB-specific development?
- Expandable trees (that we are using) only work for transitive relationships; anatomy ontology has other relationship types
- GO team does want Amigo to apply to other ontologies; how can we best coordinate with Amigo development?
- Interface with a GO-annotation verifier?
- Karen sent 280 molecules to CHEBI that we have curated with but don't yet have a CHEBI ID
- Updating can happen weekly