LEGO May 23, 2016

From GO Wiki
Jump to: navigation, search

Bluejeans

https://bluejeans.com/969313231

Agenda

Software Updates

  1. Form Interface for Curation
  2. RO View for Curators

GAF/GPAD

Still needed before we can import (into MGI):

Noctua Requirements

  1. Stable purl from which to retrieve files. (GH Noctua issue #200 - https://github.com/geneontology/noctua/issues/200)
  2. Folded annotations for individuals annotated by the regulates relation. (GH Noctua issue #189 - https://github.com/geneontology/noctua/issues/189)
  3. Strategy to create regulation target annotations in LEGO
  4. Annotation of isoforms. Can we start loading GPI files into the NEO.(https://github.com/geneontology/noctua/issues/278)
  5. Attribution to individual groups. For example 'MGI-Noctua'. (GH Noctua issue #84 - https://github.com/geneontology/noctua/issues/84)
  6. An automated QC check on annotations made in Noctua. For example, tagging annotations that don't have evidence. (https://github.com/geneontology/noctua/issues/255)
  7. Attribution of curator comments and evidence sentences to the appropriate reference. (https://github.com/geneontology/noctua/issues/280)
  8. Classic qualifiers in Noctua. (https://github.com/geneontology/noctua/issues/310)

Annotation Data flow outstanding Requirements

  1. Once we have imported annotations into MODs from Noctua, those annotations will be incorporated into output files, GPAD and GAF. Is there a strategy for them not being duplicated when they are picked up by the GOC?
  2. Will annotations in AMIGO2 continue to only come from the 'official providers' or will they be loaded directly?
  3. If annotations cycle through the 'official providers', then the official providers should all be encouraged to load GPAD files so that the associated annotation properties in column 12 of the GPAD file are not lost. How important is this since we have the original OWL saved?
  4. What is the order of priority of loaded annotations? When we strip duplicates, which annotations get priority?

Annotations/Model Discussion

Discussion of the Wnt signaling model

http://noctua.berkeleybop.org/editor/graph/gomodel:5716c41300000082

  1. Explanation of the models and the relations
  2. How will the individuals in the model chain?
  3. What annotations do we expect to get from a model like this?
  4. Is there a mechanism for limiting the annotations that are generated?
  5. How would the model differ if we had the term 'cell-cell signaling by wnt'. Draft term:

cell-cell signaling by wnt: "Any process that mediates the transfer of information from one cell to another, mediated by a wnt-family protein ligand. This process includes wnt signal transduction in the receiving cell, release of wnt ligand from a secreting cell as well as any processes that actively facilitate wnt transport and presentation to receptor on the receiving cell."

 <-- part_of wnt signaling
 <-- part_of wnt protein secretion

Question: should cell-cell signaling terms be extended to include ligand processing steps (e.g. the palmitoylation step in the above model)?

Protein Binding Annotations in Noctua

http://noctua.berkeleybop.org/editor/graph/gomodel:5716c41300000664

  • Use has input relation for annoton? has direct input?
    • In GAF/GPAD output, this will result in nothing in the With/From column
  • Use With/From column only?
  • Use both?

Expanding ECO use to Model Pathways More Clearly

  • For constructing pathways in LEGO, it could be useful to have more granular IGI evidence codes to support the assertion that gene products act in the same or different pathways
  • Work from the BioGrid, Princeton, and WB groups on a genetic interaction ontology might help inform this

Minutes

  • On call:
    • Dan (OHSU)
    • David H (MGI)
    • David OS (EBI)
    • Giulia (FB)
    • Heiko (LBL)
    • Helen (FB)
    • Kimberly (WB)
    • Melanie (EBI)
    • Sabrina (Zfin)
    • Seth (LBL)
    • Val (PomBase)

Software Updates

  • Form interface - not highest priority
  • Browse RO in AmiGO - also not highest priority
    • Not currently being loaded in AmiGO as an ontology
    • RO is available in Ontology Lookup Service (OLS)

GAF/GPAD

  • Need stable URL where groups can grab files for upload to their MOD
  • Folded annotations for regulates annotations
    • For example, curator builds a LEGO model and indicates that an activity that is part of one process that regulates a second process
    • This isn't happening yet in the GAF unless a curator makes the direct connection between the activity and the regulation BP
    • See Fgfr1 example
    • Fgfr1 activity is part of FGF signaling pathway and FGF signaling pathway regulates cell proliferation, then Fgfr1 activity regulates cell proliferation
    • In GPAD, gene products are associated with BP using the involved in relation
    • Need a logician to look at the relations to make sure the inference to the regulation process can be inferred properly
    • The inferences from Jenkins are not always been picked up by the MODs - this needs discussion on the annotation conference call
    • AI: Bring up the Jenkins inference files on an annotation call.
  • Use of has regulation target relation in LEGO
    • Need to sort this out as this relation is not available in LEGO
    • Discuss whether or not this relation should be in LEGO or if we'll need to use the unfolded relation
  • Isoform Annotations
    • In the short term, need to turn off ID checker and curators can put in any ID
    • Concern that this will be a free-for-all and groups will have to check for validity when they import the annotations
    • Don't currently have a pipeline for keeping NEO (Noctua Entity Ontology) up-to-date
    • NEO can help resolve IDs into labels
    • One possible pipeline could be to have the MODs supply gpi files to be used to populate NEO
    • NEO may also need to have information about the type of entity??
    • NEO allows for loose interpretation of entity type, but is this problematic if people want to know what the entity really represents?
  • Attribution to Individual Groups
    • GAF now specifies contributing group as GO_Noctua
    • This means that the group/curator who actually contributed the annotation is not recorded in the GAF
    • Annotations from a seeded model will still have GO_Noctua as contributing group
    • This is addressed somewhat in GPAD because GPAD includes the ORCid for the curator who made the annotation - but note that this doesn't map the curator to the group
    • Should groups just consume GPAD? What is the point of diminishing returns for trying to put OWL-based annotations into a legacy format like GAF?
    • Have groups considered how their pipelines will work to support their current database backends? Software group could help generate common tools to help with this.
    • Question arises of whether we can ever ultimately use OWL only? There will probably always be some need for a simple, table-based list of annotations.
  • Evidence
    • Multiple evidence on a relation will need to generate separate lines of conventional annotation - right now only one paper evidence, for example, is being included
    • Evidence statements from papers will need to be associates with specific papers; reworking of evidence model will allow this
  • Use of Qualifiers
    • How will these be used in LEGO?
    • contributes to may be important for annotating protein complexes
    • NOT may be more problematic
    • colocalizes with may be addressed by modeling where a gene product is active
  • Dataflow
    • Duplicate annotations - how will these be handled?
    • Will annotations from contributing group be the only source for annotations or will LEGO annotations come directly from Noctua?
    • LEGO annotations, as it stands now, are part of a separate load of annotations - see GO Models on tomadachi
    • Annotations in table view would still come from contributing group
    • Different load pipelines are firewalled from each other - separate domains
    • AmiGO is ready to go with public display of LEGO models

No call next week as Monday is a holiday in the US and UK