WormBase-Caltech Weekly Calls July 2015

July 2, 2015

Discussion Topics from IWM

• Explaining job posting options via the forum in new Worm Breeder's Gazette article
• Display of CRISPR data
  • Alleles with multiple lesions (one name, many mutations), need to be curated and mapped
• Ontology term enrichment analysis, using ontologies other than gene ontology
  • Discussed on GO call yesterday; we can/should follow up with Paul Thomas
  • Would be good to have a single central tool/resource for enrichment analysis
  • PantherDB vs DAVID

WormBook Chapters

• Paul S will review over next week and will provide feedback

Outreach

• Sending out e-mails to all labs/PIs reminding about new data forms
• Could also do more personalized outreach to a smaller subset of PI's/labs
  • Could focus on PIs not at the IWM

Anatomy

• Embryonic development, cell division timing
  • Sulston timing
  • Waterson datasets
  • Zhao cell lineage timing datasets
  • Bao lab?
• New EM reconstructions from David Hall, Scott Emmons, etc.
• Neuronal connectivity, collaborative database with Scott Emmons and colleagues

Citace upload

• Curators submit data to Wen on Tuesday, July 28th

Taking over Gene Orienteer

• Xiaodong and Sibyl working on

RNASeq data

• Gary Williams only using high quality data, taking care of all curation (including meta data)
• Public archive of rejected datasets?

WormGuides

• Bill Mohler et al working on desktop application

July 9, 2015

Expression Pattern

• Certain/uncertain qualifiers not annotated before some date
• ~3,000 ?Expr_pattern objects without that annotation/tag
• Daniela work on bringing up to date, hopefully won't take long

Expression Clusters to Anatomy & Life Stage annotations

• Many large scale datasets with tissue-specific expression data
• Much of what is in SPELL is not annotated to ?Anatomy or ?Life_stage terms
• A goal: make expression data queryable via ?Anatomy terms/pages
• Wen will make the model change proposal
• We may not want to show explicitly in widget
• There is a need for a condensed display of expression data (per gene)
• Some datasets, like the EPIC data, explicitly mention each embryonic cell name
• Need for a condensed ontology browser per gene/anatomy and gene/life stage

Proteomic analysis

• Encyclopedia of Proteomic Dynamics, contacted Wen to share data
• Wen will meet/discuss with group soon to determine what the goals are
• It isn't clear what format the data has
• Should include Gary Williams on discussions as he already processes Mass Spectrometry data

External Databases

• To what extent can we take care of the data and display of other lab/publication databases
• Many authors want to share and make links to their database/website via WormBase
• What is the best way to handle large scale dataset sharing requests that don't necessarily (for the time being) fit our data model
• We can take advantage of the "External Links" display on WBPaper pages to link out the the external databases affiliated with the paper, including a link to our FTP site with shared data files, maybe?
• At least a stop gap measure until we can properly model the data

Cis-regulatory site nomenclature?

• Barbara Meyer's lab published many "rex" (Recruitment Elements on X) sites, numbered sequentially
• Tim Schedl wondering about others' thoughts/opinions on how to, possibly, standardize the names of cis-regulatory elements
• Could be like gene names, without dash, e.g. "rex1", "rex2"
• We may want to try "WBsf-" prefix, on all element names like "WBsf-rex1", although may be only used in-house

Phenotypes

• Were there any conclusions about phenotype lookup from the Allele-Phenotype form?
• Chris spoke with Harald Hutter and others at the meeting about how to improve the lookup for phenotypes
• Would be good to provide an explicit option to see phenotypes of related (or allele-affiliated) genes, perhaps by shared GO-term annotation
• Need to think more on how to best compress display of phenotypes on gene pages as well
• We do already provide links to the Variation and Gene pages (with Phenotypes displayed) in the term information box of the form

July 16, 2015

Anatomy term page expression

• Raymond and Juancarlos are working on a display of genes that may be exclusively expressed in that anatomy object

Construct/Transgene curation

• Karen trying to make the curation of constructs & transgenes easier
• May consider merging the transgene and construct OA's
• Possibly add a construct/transgene request functionality in other OA's
  • Would those need multiple input fields?
  • Karen would take care of the details

Molecule model

• Exogenous/endogenous tags issue
• Scraping data from external chemical databases versus adding biologically relevant data from papers
• We pull data from, e.g. CHEBI, but not all molecules fall under their purview, e.g. proteins

Micropublication

• Promotion and outreach
• Micropublications discoverable in PubMed?
• Publisher = WormBase? Caltech?
• Minimal standards for publication?

July 23, 2015

Worm model for autism

• Would want to take human variations implicated in autism; look for orthologous genes in C. elegans/nematodes and find/make synonymous mutations
• Prioritize based on worm phenotypes
• Generally applies to human disease variants

Database Migration

• Thomas Down leaving WormBase in September
• Moving ahead with Datomic
• Good starting use-case for Datomic is querying Datomic-version of GeneACE
• Need to make sure documentation for migration to Datomic is available and comprehensible
• Point-people at each site: Sibyl @ OICR, Juancarlos @ Caltech
• Now need to work out the mechanics of curating into Datomic

WormBase ParaSite

• Reciprocal searches (WB <-> PS) are working well

Microarray datasets & modSeek

• Some earlier datasets were re-processed (log-transformed, or re-annotated into original replicates instead of averaged results)
• Need to try out different methods of processing raw-data (WB usually only takes in processed data)
• One pipeline can feed data into SPELL and modSeek
• It's difficult to establish/determine gold standards for assessing process performance

WormBook chapter reviewers

• Send reviewer suggestions to Paul ASAP

C. elegans proteome in UniProt

• Not a complete correspondence between WormBase and UniProt
• Cases: UniProt has entry for a protein that differs by one or two amino acids from WormBase
  • Made from translations of what cDNAs etc. have been submitted
  • Partial data, e.g. partial cDNAs translated
• Anything we can do to achieve greater consistency?
• Protein data sets are important
• Hinxton can use disrepancies as a flag to check on the gene/protein models
• Would be good to have more reciprocal linkage between UniProt and WormBase
• AVR-15, UniProt have two additional entries compared to Wormbase, differing in only 1 or 2 amino acids
• Should we pick up different entries from UniProt and store/display the data; how to reconcile?
• Possible use case: enter a UniProt ID into the BLAST/BLAT tool to identify WormBase matches

Gene Orienteer Data

• Sibyl and Xiaodong looking at data and scripts from Gene Orienteer

Precanned queries for exclusive expression

• Raymond & Juancarlos working on final details
• Intent is to display genes that may be specifically/exclusively expressed in e.g. an anatomy term

Embryonic developmental timing

• Sulston, Murray timing data sets for wild type embryonic cell division timing
• Mutant data sets are coming in as well

Genetic Interaction Ontology (GIO)

• Latest version of the GIO complete
• Juancarlos and Chris built a "genetic interaction calculator" to determine interaction types from quantitative phenotype inequalities
  • http://mangolassi.caltech.edu/~azurebrd/cgi-bin/forms/gi_calculator.cgi
• Sending out to other MODs, etc.
• Seems that although there is buy in conceptually, most curators can't afford the time for such detailed curation

Phenotype (ontology) display

• Problems with display of phenotypes (and other annotations) on WormBase, as pointed out by several people at the IWM
• Karen would like to start creating allele concise descriptions
• We need compact, intelligently ordered annotation lists, not just alphabetical lists of ontology annotations
• It would be good to show ancestors for relatedness and order
• Chris working on Python script to display all annotations in the context of the entire ontology
• We will need to see if this approach is feasible/beneficial

PATO-style EQ (Entitiy-Quality) phenotype annotations

• It is clear that some phenotype annotations require details, e.g. "drug sensitivity" annotations should have the drug involved
  • This drug/molecule annotation should be present in the details if not directly in the term itself
• Raises the issue of a number of cases where we need PATO-style EQ annotations, not just explicit phenotype terms for all possible scenarios
• This would be helpful in annotating embryonic timing and identity phenotype datasets

July 30, 2015

Wen Chen helped Wen Chen

• Wen Chen (lab) has list of genes to analyze
• Wen Chen (WB) helped process the list
• Would be good to have a simple CGI to process a list of genes in a variety of ways
  • http://tazendra.caltech.edu/~azurebrd/cgi-bin/forms/fraqmine.cgi
  • For GeneTissueLifeStage and GeneConciseDescription more datatypes easily slotted in if curator makes a file
• Is this redundant with WormMine?
  • Not for data that doesn't exist (in WormMine) yet; more agile: could be up and running within a matter of days

Interconnections between WormBase and FlyBase

• We could create more inter-connectivity between the two databases
• Sharing concise descriptions of genes
• Would be good for FlyBase and WB curators (Xiaodong?) to talk about where the links should exist at each site