WormBase-Caltech Weekly Calls March 2015
From WormBaseWikiJump to navigationJump to search
- 1 March 2015
- 1.1 March 5, 2015
- 1.2 March 12, 2015
- 1.3 March 19, 2015
- 1.4 March 26, 2015
March 5, 2015
New curation topic for WS249
- We will do Lipid Metabolism for WS249
Paper ID converter tool
- Juancarlos built a simple CGI to convert a list of PubMed IDs to WBPaper IDs
- Would other curators use this or a more comprehensive tool?
- What ID types would curators want to convert between?
GO data model change
- Change for WS248
- More detail with each annotation
- Pipeline for data incorporation into the build has changed; path goes through Caltech/Postgres for QC
- Changes to gene association file has implications for automated descriptions
- Do we want to include PAINT annotations for automated descriptions?
- Based on phylogenetic evidence codes
- Yes, could be a great benefit to the automated descriptions
- New GO annotations include a Panther tree node
- Should we include both manually annotated data AND PAINT data in automated descriptions equally? Should look at examples and check how we want to handle
- Clickable (linkable) evidence throughout the description?
International meeting give-aways
- Worm pick holders (pencil grips) with WormBase logos; can they be printed on?
- T-shirts and/or hats/caps; personalized for curators
Topic & Regional meeting abstracts
- Usually we receive the data for these meeting abstracts in a non-standardized format; makes it difficult/time-consuming to parse
- Would be good to request submissions in standard format
- Good to maybe go through GSA pipeline?
NIH request for info on sustaining bio-databases
- NIH requesting info on:
- best practices
- alternative funding sources
- requiring a DB funding allocation in individual grant applications?
- medical application
- What benefit do databases provide?
- Differential usefulness: transient DBs versus maintained, broader impact DBs
- Should MODs merge?
- So much has changed in terms of database and web technology since online bio DBs began
- Value distinction of manually curated annotations
- How much redundancy is there between DB effort?
- How do we assess the value of WormBase (or any DB) for academia and industry?
- We should assess the scientific value to determine monetary value
- Promote the benefit of transparency and reproducibility
- We can introduce the idea at the international meeting
- May be nice to group all micropublications into a single official publication
- Maybe use "Gene Expression" journal for gene expression data as a pilot
March 12, 2015
- Issues came up with layered hash references in ACEDB when converting to Datomic
- Changing line from "Historical_gene ?Gene #Evidence" to "Historical_gene ?Gene Text"
- Affects ?Interaction, ?Expr_pattern, ?Antibody, ?Construct, and ?WBProcess
- Do we want/need a separate chapter/section on the new WormBase Ontology Browser?
- Maybe only need a few small paragraphs in other relevant chapters, should be sufficient
- Gene Function & Interaction chapter almost complete
- Need to finish intro for phenotypes
- Need to add predicted & regulatory interaction descriptions
Topic OA reminder
- The 'primary'/'nonprimary' paper status is updated automatically every night by a cron job
- The "Curation Status Omit" toggle button on TAB 3 was intended to flag nonprimary papers, but is no longer needed for that; only for rare noncuratable, primary, relevant articles (e.g. Methods papers)
Cloud bio-data storage
- Cloud has potential advantage of distributed colocation; robust for queries, performance etc.
Modeling user-submitted data
- How should we distinguish user-submitted data from curator-submitted data?
- Should we distinguish at the data model level, with a new tag for example? Probably only need to apply WBPerson evidence, etc.
- This issue came up with user-submitted/user-edited concise descriptions
- The distinction can be made explicit exclusively at the display (web page) level
- There could be a toggle in the OA to indicate user-submitted data
March 19, 2015
- Most chapters mostly complete as a draft
- Paul S. working on introduction
- Process chapter may not fill a chapter; also pending some website updates
- Person/author section can be added to nomenclature section
Mapping Bio-Annotations to Process
- Can we achieve automated mapping of annotations to processes?
- e.g. phenotype-to-process mapping, site-of-action (cells and genes) mapping to process
- Some phenotypes are indicative of a process, but some cannot be unambiguously mapped
- Some processes can be mapped to based on sequence homology (e.g. Wnt signaling) rather than phenotype
GO Physical interactions into WormBase
- Several (~400) papers in curation status form are considered "curated" for physical interactions via the GO pipeline
- We would like to import these annotations into WormBase
- Directionality information has been lost in the GO annotations
- We want to re-curate the interactions to capture detection method and directionality
- We could import all interactions in bulk without detection method or directionality with flag indicating a need to re-curate with additional information
- Reminder for curators to check their lipid metabolism papers
- Daniela working with web team to create a Pictures widget on Topic pages to display pathway/topic images (with permission)
- On staging, Sibyl has added an "Interaction-based Phenotypes" table in the Gene page Phenotypes widget
- This phenotype data is taken exclusively from genetic interaction objects to associate the gene with the phenotype
March 26, 2015
Community submission forms
- People may submit expression data directly, with future citation (e.g. DOI)
- Data submissions will need to adhere to some criteria providing evidence and reviewed by Paul, Daniela, and some other reviewer
- Form could be recycled for published data submission
- Forms remember user e-mails from IP used; makes easier to fill out subsequent submissions
- Daniela will talk to Daniel about assigning DOIs
- Can these be indexed by PubMed, etc.? Daniela discussing with Caltech librarians
- We should focus on promoting/pushing one data type submission so as not to overwhelm the community; focus on others later
- Concise description forms
- Want to send automated descriptions (or existing manual descriptions) to users to quality check based on their expert (published) knowledge
- User will use form to submit edits to descriptions
- Do we want/need users to stick to automated description format/guidelines?
- We could have direct links to forms from the Overview widget on gene pages
Funding Supplement Dedicated to Interoperability between MODs
- Textpresso can be used to create common text mining pipelines across databases
- Combining concise descriptions or taking advantage of other MOD data
- Supplement funding application due in April (soon)
- Hope to have a pilot of new databases (Datomic?) done by mid-Fall
Global correlation of RNAseq gene expression data
- All pairs of genes could have a correlation calculation performed (based on RNAseq) and displayed on WormBase
- Significant correlations could be a type of interaction/association