WormBase-Caltech Weekly Calls June 2017

From WormBaseWiki
Jump to navigationJump to search

June 1, 2017

Upload

  • After worm meeting, (June 27th)

IWM swag

  • Rubber bracelets?
  • Left over bandanas? paid ~$1 per bandana before. We have enough for WormBase staff
  • Water bottles?
  • Small boxes of mints?
  • Will need to decide soon to have enough time for ordering

Phenotype modeling

  • Chris still working on ?Phenotype_assay (?Phenotype_experiment/?Phenotype_annotation) class model
  • May require a new OA that supersedes the Phenotype OA and RNAi OA
  • The new model still requires the #Phenotype_info hash as long as we still want to make assertions (e.g. loss-of-function, gain-of-function, recessive, dominant) about specific perturbations (e.g. variations) in the context of the phenotype experiment
  • Have tried to pull penetrance info out of the #Phenotype_info hash and into the ?Phenotype_assay model, but this would require only ever having one phenotype (or one set of phenotypes) with penetrance values per experiment
  • There are several ?RNAi objects in ACEDB that refer to multiple phenotypes, each with their own remark and penetrance values. Question is whether we want to separate out all the individual phenotype observations into multiple distinct ?RNAi objects and, soon, ?Phenotype_assay objects
  • If the only time we want to apply a value like "gain-of-function" or "recessive" to an allele is when a single allele is used, we could get away with keeping those tags in the ?Phenotype_assay model and not use the #Phenotype_info hash
  • Chris will share the model proposal and put together some examples for next week's meeting


June 8, 2017

IWM swag

  • Google Doc
  • We have left over tattoos, fly swatters, bandanas
  • Something to identify WormBase staff? bandanas, lanyards, hats, glow sticks/necklaces?

?Phenotype_experiment data model

  • New ?Phenotype_experiment data model proposal HERE
  • ?RNAi could still be used going forward to represent the RNAi dsRNA sequence or reagent used to knockdown gene expression
    • Then RNAi sequence information would not need to be stored in the new ?Phenotype_experiment object
  • The phenotype connections would change in the data model
    • Phenotype would no longer be directly connected to ?Variation, ?Transgene, ?RNAi, or ?Rearrangement
    • Phenotypes would be associated with these objects indirectly via the ?Phenotype_experiment object
  • Perturbed_gene tag could be populated automatically in most cases (from ?Variation, ?RNAi, and ?Rearrangement objects) and some cases manually when, for example, authors don't specify a mutation but just say something like "xyz-1 mutants exhibit this phenotype"
  • We will change the dumping process to dump some Phenotype_info data, like variation effect (e.g. gain of function), directly into the ?Variation objects that they apply to; should make querying and displaying the data easier
  • Are we trying to capture too much detail? No new detail is being added in this model; curators feel that the detail is not the rate limiting step of curation
  • Querying out (or displaying) which Phenotype_info goes with which paper is currently difficult/problematic; for example the e1370 allele is annotated to the "dauer constitutive" phenotype about 20 times, but the details are all concatenated in the web display, making it impossible to tell which remark (e.g. "Table 3") goes with which paper
    • Although the current data model doesn't make this type of distinction impossible, it is cumbersome and would likely benefit from a ?Phenotype_experiment class
  • How do we handle cases when multiple genes are mapped to a ?Rearrangement object or ?Variation object? How do we display that multiple independent variables (e.g. alleles) are responsible for a phenotype?
    • Possibility is to have phenotype table display an "Independent Variable(s)" column that displays all of the elements that likely contribute to the phenotype; at least then users are aware that multiple factors could be playing a role

Open position for WormBase

  • Has been posted for a little while now
  • Send any suggestions to Paul
  • Could announce at the IWM

Phenopackets

  • Karen will send around info on Monarch phenopackets
  • They've requested contact with a WB curator


June 15, 2017

IWM Plenary talk

  • Paul can mention SObA: need a good example to demonstrate the power of SObA
  • Will also discuss AGR, community curation, micropublications, data mining tools

Micropublications

  • Daniela and Karen have been working on booklet
    • Will print ~500 copies
  • Will be raffle for those who try the micropublication form ($20 voucher per day)
  • Daniela finalizing poster
  • Have booth next to WormBase

IWM swag

  • There will be glow sticks
  • Karen ordered bling lanyards for the micropublication project
    • Will be limited red ones for WB staff if they want to help advertise, they do not say WormBase on them though.
  • WormBase tattoos; Paul left them with Ranjana (Ranjana will bring to the meeting)
    • Karen also has some from the TAGC meeting

IWM Workshop

  • 7 minutes per talk? Should leave enough time for transitions and discussion (at the end)
  • Ranjana will send email around to the group to update on the timing

?Phenotype_experiment model

  • Removed RNAi sequence specific tags, will leave in ?RNAi class so ?RNAi objects act as sequence reagents (like alleles)
  • Also removed variation effect tags etc. to dump explicitly in the ?Variation etc. class

Phenopackets

  • Chris joining call tomorrow with Monarch, Nematmetrix, and Knudra about WB phenotype data and phenopackets
  • Chris will relay the conversation to the rest of the group
  • Probably won't need to change curation or data model, just need to dump data in another format

Chat service for IWM communciation

  • WhatsApp group?
  • We can probably just use Slack; will allow pop-up notifications on phone


June 29, 2017

Wiki login problems

  • Several curators are having trouble logging in to the WormBase Wiki
  • Message says cookies need to be enabled, even though cookies are enabled/haven’t made changes since last login
  • Chris (after meeting): Seems to be resolved now

WS259 delayed

  • WOBr and related tools are using/displaying WS259 data even though the rest of the site is WS258; some asynchrony

IWM

  • Raymond: lots of neuro data
    • Serial reconstruction
    • Activity data
    • Neural pathway data
    • Worm tracking systems
    • Modeling
    • Finding receptors/ligands for orphan ligands/receptors
  • Wen
    • Received questions about genomic data, SPELL, RNAseq
    • People were surprised to learn about SPELL and ability to download data sets
  • Ranjana
    • WB workshop was well attended
    • Users had a few questions about the tools
    • Lots of interest in micropublications
  • Kimberly
    • People are using GO
    • People find the ontology very large and intimidating
    • Hopefully Noctua can help replace long, complex terms
  • Chris
    • Denis Dupuy presented a new method for displaying prevalence of transcript isoforms
    • Would be good for WB to represent splice variants on a log-scale of prevalence using Denis’ program
    • Denis is already in discussion with Hinxton team to try to implement

SPELL

  • SPELL is still not supported/maintained; no replacement in the works
  • ModSeek is finished; student graduated, no longer supported
  • SPELL data is growing, fewer microarrays more RNAseq
  • We could mention that data sets are downloadable from SPELL on WB blog/news post; recur monthly/yearly
  • SPELL sometimes freezes; probably people performing large queries; could block additional users from accessing SPELL
  • Wen needs to remove log files every few months to make space on the server
  • Would AGR be able to support SPELL or similar tool?

Retracted papers

  • Should be available but clearly labeled in Textpresso
  • Curated data from retracted papers should be omitted from WB