WormBase-Caltech Weekly Calls December 2014

From WormBaseWiki
Jump to navigationJump to search

December 2014

December 4, 2014

Institution tag in ?Person class

  • Cecilia gets data from papers (affiliation) and direct person submission.
  • Some people are affiliated to several departments or several institutions
  • Do we want to capture and distinguish all affiliations?
  • Should the info be collected and pipe separated inside the Institution tag?
  • This potentially adds a great deal of work to collect, fix, and maintain this info
  • Do we want to create an ?Institution class? Do we want an ?Affiliation class?
  • We want to standardize country name and institution name (use United Nations for reference of country names http://unstats.un.org/unsd/methods/m49/m49alpha.htm )
  • Institution and department can be listed in the same line like
?Person
 Institution ?Text Department ?Text
    • Or if keeping as
 Institution ?Text
    • We'd tokenize the text with pipes or semicolons, separating the Institution name; Insitution city / state / country; Department
    • Or we'll put the Departments in the Street_address tag instead of the Institution tag, making the Insitution tag a controlled vocabulary
  • Standardized county and state names?


December 11, 2014

Updating Constructs

  • Constructs previously only with a text description are getting updated to fit the data model
  • Relevant genes and reporters are being moved to the appropriate tags in (where possible) controlled vocabulary

Interaction-Gene associations

  • Variations should be pushed from #Interactor_info into Interactor tag of ?Interaction model
  • Variations could be mapped to genes during the build
  • Transgenes can still be in #Interactor_info but relevant genes (from the transgene) will need to be annotated explicitly
  • Hinxton (via the build process) will need to fill in the gene interactor for interactions where variations are annotated as the interactor
  • Karen and Chris will work with Juancarlos on how to handle dead genes in ?Construct models and mentioned in interactions via transgenes

WormBase Ontology Browser

  • Updated to WS246 on the local dev server

Expression Clusters

  • Large increase in numbers of papers/datasets with expression cluster data
  • Datasets including RNASeq and proteomics

Recap on Todd's visit

  • Updating interaction display
  • Separating Interaction-based phenotypes from directly caused phenotypes
  • Added sequence features as interactors in Cytoscape interaction view
  • Updated GBrowse view of sequence features
  • Institutions - affiliating authors to institutions and countries with controlled vocabulary

Single molecule FISH annotation

  • The number of molecules observed is captured in free text

High throughput variation annotation

  • Karen and Mary Ann are curating large numbers of lethal (Let) variations being sequenced by Ann Rose and David Baille
  • Mary Ann has curated the molecular lesion information for these alleles - they'll be in WS247

Pathway diagrams and images from papers linked to Topics

  • We will begin annotating paper/review pathway images (with permissions) to WB topics


December 18, 2014

Topic Diagrams

  • Working out a pipeline to capture pathway diagrams from papers
  • For pictures we have permissions for, we will display the images on the Topic pages
  • Topic OA has two fields added to flag papers with good diagrams (toggle field) and a text field for curators to indicate the figure(s) number(s)
  • When curators scan through papers from a new topic for relevance, they can also flag a paper for good diagrams
  • We may be able to import standard format images from PubMed Central

Personal Communication

  • Do we want to take any unpublished expression data from labs/PIs?
  • We could, but the contributors should fill out a form to submit the data
  • This could be a good way to test out community annotation forms
  • We want to establish a micro-publication pipeline whereby data can be submitted formally with minimally required information (enough to reproduce)
  • We need to be very clear on the data display on the web as to whether or not the data is personal communication (if not a formal micro-publication)
  • Applies to all data types; common examples include allele-phenotype data
  • Need a disclaimer statement like: "Unpublished data, cite only with author permission"