LEGO October 24, 2016

Bluejeans URL - NOTE NEW MEETING TIME: 8:00am PST

https://bluejeans.com/969313231

Agenda

Software Updates

GAF/GPAD Updates

• Still working on this with Wednesday group
• Not all traditional GO annotations are being generated in the GAF/GPAD output

Models Review

• Question about updates in RO
  • We had talked about having positively and negatively causally upstream of relations
  • Are these in the RO?
• GPCR template
• RGD annotation consistency exercise
• Relating metabolic processes to organismal level processes - see github ticket #25
• Other templates that might be helpful
  • Signaling pathway modules - see github ticket #2
  • Cellular processes
    • Vesicular trafficking
  • Metabolic pathways

Minutes

• On call: Barbara, Chris, David H., Elena, Giulia, Helen, Jim, Kimberly, Midori, Moni, Paul T., Ruth, Sabrina, Seth, Stacia

Software Updates

• Jim and Seth wrapping up some fixes, will alert curators
• These fixes will address the Monday morning model issue
• Also, a new login system will be coming along, so curators/users will be asked for either a github account or Google ID

GAF/GPAD Updates

• David H, David OS, Jim, Kimberly meeting on Wednesdays
• Jim working with GO graph store which holds all of the models
• Queries over the inferences will generate the GAF/GPAD annotations
• Lots of inferences generated, but then pruning will happen for output files
• Only production models will be loaded into the file-generating pipeline
• A slightly different, but more dynamic, system will work for more real-time file output while using the tool
• Most of the work will be handed off to the SPARQL query, rather than the existing minerva code
• Evidence issue worked out? Not yet.
  • Current minerva code has evidence on the asserted stuff, not the inferences.
  • In the future, evidence will be structured differently in the models.
  • e.g., the PubMed will not be a literal, but a URI

Models Review

• Relating metabolic processes to organismal level processes - see github ticket #25
  • Broad question - how to relate causal chains where one process may be causally upstream of another process
  • In these cases, what evidence should be used for relating the two processes?
  • There is always the question of whether we should make these annotations at all, but if they exist in the models, we agree that representing the causal chain is most biologically intuitive and the evidence on the edge relating the two processes should be used to support annotations to the latter term
  • We recognize that with evidence codes like IMP, curators may be using more information than just the mutant phenotype to infer the relation
• Updates in RO
  • Adding positively and negatively causally upstream of - Chris will get these into RO, Seth will get them into Noctua
• Review of GPCR template
  • Work on this template came from the dictyBase annotation consistency exercise on cAMP signaling
  • The idea is to try to capture the minimal amount of information for defining a pathway to then use as a template
• Signaling templates
  • See github ticket for list of possible signaling templates
  • If you have interest or expertise in one of these areas, please contribute a template
• Use of PANTHER class to indicate the type of molecule expected to perform the activity
  • Paul will follow up with this
  • We'd also get built in taxon constraints
• RGD Annotation Consistency Exercise Model
  • Iron absorption in the rat intestinal epithelium
  • Look at transport of ferric ion from the intestinal lumen to the blood vessels
  • Assess the transport with antibodies to the different components, e.g. transferrin, transferrin receptors, ferritin
  • Relatively straightforward, but there is a question about how best to represent the relation between transferrin and transferrin receptor
• Templates
  • Also would like a template for gene silencing by miRNA
  • Generally, would be good to also have templates for various types of regulation of gene expression