WormBase-Caltech Weekly Calls January 2019

From WormBaseWiki
Jump to navigationJump to search

January 3, 2019

WS270 Citace upload

  • Next Tuesday, Jan 8th, 10am Pacific

Gene descriptions

  • Valerio generated new files to ignore/filter-out problematic genes
  • Still need to validate new pipeline
  • Barring any major issues, will submit new files for WS270 (can load old files if needed)
  • Maybe should define a test set (random sample) to test each release? Already have a test set

Protege Tutorial


January 10th, 2019

WB workshop at IWM 2019

Here's a draft, need to finalize as Jan 15th is the deadline


Title: WormBase 2019 - Data, Tools and Community Curation
This workshop will be an interactive session with users in order to discuss the types of data in WormBase and how to query them using specific tools.  We will discuss recent changes to WormBase community annotation forms and how to use them to contribute data to WormBase.  We will also present updates to ParaSite, a portal for parasitic worm genomic data, and guide participants on how to find data across model organisms at the Alliance of Genome Research.

Format: 90 minutes: 1 section of 40 minutes, followed by a second section of 20 mins and a third section which will be a 30 minute open discussion/Q&A session.  Talks in each section will also be tailored to allow time for questions from the audience.

Section 1: Introduction to the WormBase gene page and tools such as SimpleMine, Tools for RNA seq data and enrichment analysis, gene-related data using WormBase Ontology Browser and Annotation Visualization tools.

Section 2: WormBase Parasite database, Model Organism data at the Alliance of Genome Research and Community Curation forms

Section 3: Open forum for discussion and Q&A.

Finalize Protege tutorial time

  • Best final options:
    • Wed, Jan 16th, 1pm Pacific/4pm Eastern
    • Thurs, Jan 17th, 11am Pacific/2pm Eastern
    • Thurs, Jan 17th, 1pm Pacific/4pm Eastern
  • Propose we go with Wed, Jan 16th, 1pm Pacific/4pm Eastern

Automated descriptions

  • Distinguishing information rich vs. poor genes
  • Information poor genes can take advantage of information across MODs/species
  • Need more robust QC pipeline; can work on for WormBase, and later apply to Alliance once worked out
  • Working on expression statements for Alliance genes
  • Considering rearrangement of description so disease features more prominently

Disease curation

  • Disease model curation progressing; Lots of discussions about data standards and entities in Alliance Disease Working Group
  • Considering SVM for disease; current paper flagging pipeline is rather broad
  • 200+ papers as positive training set available
  • Results section are not being extracted in latest Textpresso (paper sectioning in general not happening)

Noctua / GO-CAM

  • Making progress on best practices
  • Can use Noctua to generate GO annotations
  • Starting to incorporate proteins
  • Working with an ever changing Noctua platform; bugs emerge as it is developed; may benefit from frozen release of the software
  • Next month or two, will import entire set of C. elegans GO annotations into Noctua
    • Many decisions to make: how to model?
    • Each gene will become a single Noctua model; not linked to each other initially
    • Working on batch updates/uploads to Noctua

Expression cluster curation

  • Wen working on 40 paper backlog; hoping to finish by WS271
  • Wen wants to work on RNA-Seq tools next
    • FPKM tools
    • Filtering by datasets
    • Would like tools ready before International C. elegans Meeting (June 2019)

Neural function curation

  • Raymond: want to use design pattern strategy to curate

WOBr

  • Now incororating non-IEA disease annotations into WOBr
  • Using disease-association file

Phenotype curation

  • Will run a new round of phenotype requests on ~3,000 papers in next few weeks (last one ran in October)
  • Processing community curation submissions
  • Will recurate some community curation papers to check:
    • 1. completeness of community curation
    • 2. the time-savings of the phenotype form pipeline
  • Have made recent improvements to phenotype request emails, allowing authors more feedback options which are now being readily used
  • Working with new phenotype ontology GitHub repository
    • OBO Foundry now pointing phenotype ontology at the GitHub repository (both OBO and OWL files)
    • Need to update the citace upload procedure to generate phenotype .ACE file; currently the script is still running on the old OBO Tazendra location; need to update to work off new OBO file at GitHub

Metabolomics

  • Karen working with Michael Witting to pull in metabolomics data
  • Integrating information about endogenous concentrations of metabolites

Automated descriptions React tool

  • Juancarlos developed tool to request versions of the automated descriptions
  • Will update pipeline to pull data from Alliance; currently coming from Tazendra
  • Tracking how the descriptions are changing, by data module for example
  • React tool currently on mangolassi but will move to Alliance at a location of Olin's choosing (AWS resource)

Transgenes in the Alliance

  • Are transgenes being discussed at the Alliance?
  • Yes, the phenotype and disease working group has been discussing
  • Hasn't come up in recent weeks, but was discussed at face-to-face meeting
  • One significant issue is that WormBase uniquely has extra-chromosomal arrays, whereas other MODS (always?) have integrated transgenes and consider them types of alleles
  • Chris will give Karen a heads up next time the issue is intended to be discussed within the Alliance


January 17th, 2019

Alliance Grant


January 24th, 2019

Author First Pass

  • For strain identification, we are using the obo_name_strain table.
  • There is an entry for 'Strain' in that table that leads to false positives.
  • Is this entry needed for curation?
  • If so, we will just filter it out for the purposes of AFP.
  • In Tazendra with timestamp Jan 23, 2019; on Mangolassi with timestamp Nov 15, 2018
  • In WS269 with timestamp '2018-09-25_17:00:39_pad'
  • Linked to paper WBPaper00055300; Location 'PS'; species C. elegans

Specifically expressed genes

  • On anatomy pages, in the Ontology Browser widget, we have a list of genes in a box that says "There are ### genes that may be specifically expressed."
  • These genes are genes that (1) are shown by expression pattern (Expr_pattern) objects to only be expressed in that tissue/cell or subtype but not in any other AND (2) genes that are shown to be enriched in that tissue/cell or subtype by expression cluster data BUT may include genes that are shown to be expressed (to some degree) via expression cluster in other tissues, albeit at low levels
  • Wording is currently a bit misleading; should the statement/wording change or should we change the algorithm?
  • Could offer a specifically expressed list and an enriched list separately
  • Warrants more discussion


January 31st, 2019

IWM workshop

  • Organizers reluctant to give us three workshops
  • We could ask for two: Micropublication & WormBase
  • Will be difficult to cover all relevant material
  • WormBase content
    • Would want to cover tools for retrieving data
    • Want WormMine, JBrowse, ontologies and ontology tools
    • Should consider: Paulo is working on InterMine 2.0 release which will have a different interface; we should think about the timing of that release with respect to the timing of the meeting
  • We can try to have tutorials/presentations at the booth
    • Set up a schedule to cover certain topics when appropriate curators/staff are there
  • We can ask attendees of the workshop what additional material to cover at the booth later
    • Prepare several presentations, and present those that get the most votes

Specifically expressed genes

  • Proposal to:
    • A) Parse gene sets that fall into specific categories and provide several lists
    • B) Keep as is but make the language of the statement more vague
    • C) Keep as is but remove genes shown to be expressed (to any level) in other tissues by expression cluster data
  • Could reserve "specifically expressed" for genes shown to only be expressed in that tissue by Expr_pattern data and/or expression cluster data, but excluding genes expressed in any other tissue to any level
  • Decision: Will change text to "may be predominantly expressed" and add explanation text for users
  • WormMine currently linking genes to any anatomy term (?) associated with an expression cluster; we need to review this as sometimes genes are connected because they are depleted in a tissue.