Difference between revisions of "WormBase-Caltech Weekly Calls May 2017"
From WormBaseWikiJump to navigationJump to search
(Created page with "== May 4, 2017 == === Strategies for GO-CAM (Noctua) curation === * GO-CAM (Gene Ontology - Causal Activity Model(ing)) * Proposal for curation report page (Kimberly) * Paper-...")
Latest revision as of 15:31, 1 June 2017
- 1 May 4, 2017
- 2 May 11, 2017
- 3 May 18, 2017
- 4 May 25, 2017
May 4, 2017
Strategies for GO-CAM (Noctua) curation
- GO-CAM (Gene Ontology - Causal Activity Model(ing))
- Proposal for curation report page (Kimberly)
- Paper-by-paper approach vs. big-picture approach
- Kimberly presented proposal for curation tool:
- Given a list of genes and keywords, tool could pull up all relevant annotations to help start with GO CAM model building
- Tool could return GO annotations, expression patterns, phenotypes, interaction-based phenotypes, etc.
- Could have a table to fill out that could then be imported into Noctua
- Can we use SObA to visualize relevant annotations for a set of genes? Help visualize common vs. distinct annotations?
- Process could highlight the need for micropublications
- Raymond & Juancarlos have been working on autocomplete tool for search bar; not trivial
May 11, 2017
Phenotype Strain curation (ky)
- Changing Strain tag in the Phenotype class, which is used as a primary object for annotating a phenotype to, to clarify its purpose.
- A Strain tag also exists in phenotype assay, which is meant to capture the parent strain in which a mutant or other primary object (transgene or rearrangement) is part of.
- These two tags have been causing confusion for curators and it causes errors upon the dump, in addition strains entered into the Control_strain field of the phenotype form are throwing errors during the dump as they are being recognized as "Strains".
- The dumper does not allow more than one primary phenotype affected object, it can be Variation OR Strain.
- Control_strain data from the phenotype form are being dumped as "Strain", not Control_strain. Chris will work with Juancarlos to fix
Metabolic curation (ky)
- Michael Witting has hired a student to curate elegans metabolic data.
- I will be talking to him to see what we can provide to help them share these curated data with WB. This might be as minimal as an spreadsheet although if they are submitting their own data, I will aim for a micropublication template. If there is another curator that wants to sit in on a call with Michael and me, that would be appreciated.
- Mary Ann will join the call
- CGC site now has a login. While we will be able to access all the info without logging in (the log in is required for placing an order or viewing an order history), do we need want a login?
- Aric talked with people at wormbase a while ago about the changes, was there any decision about if there will be a wormbase group login. I asked if they will be using logins to track activity on the site, he said no.
- We do not need to do anything, we have the access we need.
Methods in Molecular Biology
- Received comments back from editor, Chris going through them now
- Chris is working through the comments; will send around to staff
SObA Gene Ontology interface
- Raymond and Juancarlos have resolved most suggestions
- Have settled on a search scheme
- Haven't yet implemented highlighting GO Slim nodes; will work on
- Maybe allow user to request "highlighting" of GO Slim nodes
- Button "Show GO Slim terms"? Default highlighting of GO Slim nodes?
- URL should still be active
- There is a WormBase progress report and an AGR progress report
- Send WB reports to Paul
- Putting phenotype remarks in their own column; in-line evidence
- Adam & Sibyl are working on phenotype display from Datomic
- We can wait until they ask us to review to make comments about phenotype display
- What about changes to phenotype model? Will make changes to phenotype model soon, but display shouldn't wait
- 3 year vision request; where do we want to be in 3 years?
- ACEDB to Datomic? AGR dependent activities? May need to consider both
May 18, 2017
GO SObA update
- Auto complete complete?
- GO slim nodes highlighted, example.
- List of AGR slim terms in legend
- GO terms link out to Amigo
- Raymond, Juancarlos will remove the "Fake Root", "Filter LCA Nodes", and "Filter Longest Edges" checkboxes in legend (for simplicity)
- Will add to legend filled (or half-filled) circles to indicate slim terms
Texas Women University visit report
- Tissue enrichment tool (positive feedback)
- Want to see network for a list of genes
- WormMine and Simplemine
- ParaSite BioMart
- User's traffic
- Community curation
- For Mitani alleles:
- Suggesting one micropub per allele
- Need to reach out to Mitani lab
- Could mention at IWM
- Spoke to Michael Witting
- Interested in submitting micropubs for metabolites
- Received micropub submission from prior reviewer
- How readily can we clone existing forms (e.g. Phenotype form) for micropublications?
- Depends on what we want to do
- There need to be micropublication-specific differences, e.g. don't need to submit a paper ID
- Might be best to keep one form but modify to accommodate micropubs (so multiple forms don't need to be changed whenever there is a change)
- There is already a checkbox in the Phenotype form for "personal communication"; could add another for "micropublication"
- There was an increase in weekly visitors to the micropublication page following the blog post
IWM Plenary talk
- Paul S will give talk
- Will discuss AGR, micropublications, community curation, new data (metabolites?), human disease, connectomics?, Textpresso central
- Michael Witting will get a curator; may contribute to WormBase
- Would need to give her an OA tool; access to Postgres
- Jamboree for metabolomics modeling in November
Negative expression data
- Help desk inquiry (from Hobert lab) for negative expression data
- Student(?) may have negative data to submit (micropublication?)
- Current annotations limited to explicit statements of negative data, e.g. "Expressed in QL, not in QR"; annotate to "Not in QR"
- BGee has tool to search for presence or absence of expression based on RNAseq data
May 25, 2017
- Updated expression data (had incorrect data associations)
- Need to update expression query template
- RNAi data needs some updates
Network issues at Caltech
- No big problems so far
- Thurs - Sun
- Tuesday June 27, 10am
- Chris building up new ?Phenotype_assay class
- Later remove tags no longer needed from other classes
- RNAi objects will likely still exist as containers for RNAi sequences; act like alleles
- Ultimately may want to pull everything from the RNAi class into the ?Phenotype_assay class
- Plan to have an "Independent_variable" supertag under which we would have references to any and all entities thought to be causing or contributing to the phenotype
- Independent variables could include: genes, variations, transgenes, rearrangements, RNAi objects, molecules, strains (natural isolates), free text
- Still need to keep the #Phenotype_info hash, but can possibly add XREFs to populate nature of alleles into allele objects themselves; make querying easier for that type of data
- Will reference natural isolate strains as independent variable, separately from "parental strain" and "control strain" to capture strain carrying allele, for example, and control strain
- Condition will still be captured as free text, but we may want to switch to a controlled vocabulary at some point, in line with ZFIN and RGD
- Modeling process will require changes in phases to allow migration of build steps and website display over to new model
- Doug Howe from ZFIN asking about micropubs for zebrafish community
- Large portion of zebrafish community interested in micropubs: expr pattern, CRISPR, morpholinos, alleles, etc.
- We could survey for interest at International C. elegans Meeting