LEGO September 21, 2015

Attendance: Rama, David H., Paul T., Kimberly, Seth, Chris, Heiko

• Moving forward with LEGO-style annotation:
  • Documentation
    • Between now and end of November, we will work on documentation for:
      • Using the tool (start from existing videos?)
      • Philosophy behind LEGO-style curation (introduction to the process, how it differs from current curation paradigm)
      • Specific examples of curation
        • Single paper (maybe from an annotation consistency exercise)
        • Conserved pathway/textbook example (glycolysis)
        • Up-to-date view of a biological process (using summary figure from a recent paper or review as a starting point, annotations derived from multiple papers - sometimes this is easier for curators to model)
  • Roll-out
    • Early December? Will need to put together some possible dates and see what works
    • An in-person meeting with representatives from each group currently annotating with Column 16, and others, if funding permits

• Issues for discussion
  • Use of TAS and NAS to create reasonable biological models when evidence is either not annotated yet or not available in the literature
    • Think about how this would be visualized by users to be clear on what is known and what is proposed by authors
    • Should TAS annotations point to a particular paper in With/From column? Could the PMID in the With/From column be used to fill-in annotations in a LEGO model?
  • Cross-species and cross-tissue modeling
    • Is used in biology, but our current annotation model doesn't quite handle the representation of these types of inferences
    • New evidence code?
    • Refer to other annotations (using stable IDs) to document reasoning - inter-model evidence
    • How often does this occur? What is the pragmatic solution for handling this?
  • Highly specific vs general models
    • Curation will likely happen on the more specific models, but we can use that information to develop a more general, or common, view of a biological process