WormBase-Caltech Weekly Calls December 2011

December 1, 2011

WORM paper

• Sorted out at conference call
• EBI will take lead on rewrite
• We want a paper on the new website to go to a broader audience

Interaction model

• Almost sorted out and working
• Chris & Kimberly will discuss

Elsevier linking (images)

• Caltech attorney's want the agreement with Elsevier to be under US law
• Usage data relating to Elsevier content access must be deleted within 3 months(???)
• Should not be an issue?

Local AQL queries

• Running on Caltech mirror? Currently getting an error: "Connection closed by remote server"
• Can run AQL from xace on spica

December 8, 2011

Bio Curator Meeting

• In Washington DC next April
• Wen will send an abstract on SPELL
• Kimberly, Arun and Michael will be there as well
• Abstracts to submit: 3 in total (Kimberly, Arun, Wen)

WomrBase curator ad officially posted

WormBase meeting announcement

• What is our policy? Should we send info to everyone?
• We can send to PIs only
• Meeting announcements are an appropriate use of the list (if a bonafide C. elegans meeting)

Beta site Protein Page review

• Common protein name e.g. 'UNC-26' or 'Synaptojanin'
• Use GO curation, PAINT/PANTHER tool to assign name?
• What's the best method (automated, if possible) to assign a useful name?
• Refer to Uniprot for names; can EBI pull names directly from Uniprot?
• GO: Molecular Function and Cell Component are annotated to gene, in some cases an isoform is attributed
• Want to enable curation to isoforms when info is available

Interactions

• Interaction model still being tweaked
• Will try to 'swallow up' Gene Regulation objects and tags into main Interaction model
• #Interaction_info and #Interacor_info may be removed

Anatomy nomenclature discrepancies

• Oliver Hobert raised issue of consistency of names
• For example, sometimes "DA neurons" refers to dopamine neurons as opposed to the anatomical 'DA' neuron
• DA neurons are actually cholinergic
• Maybe post on the blog
• Make synonym for dopaminergic neuron as 'DA neuron'?

Citace upload date (WS230)?

• Monday, December 19th
• Scripts to be run manually

GSA Markup at FlyBase

• Flybase's role in markup and QC?
• 3 month notice to stop curator QC at FB
• FB curators do not want to QC, even with the 'QCFast' tool
• How to make the tool user-friendly for authors?

December 15, 2011

Interactions display on Beta

• Under 'Phenotype' reads "Interaction_Phenotype" e.g. let-60
• On interaction page, no reference

Dead genes in interaction objects

• In citace minus, dead genes hidden in hashes
• Some data stored as .ace files, others (large-scale) in flat files; flat files easier to manage
• We don't want large-scale data in the OA; take up too much space? (Megabytes of data, ~quarter million objects?)
• Interaction objects have to be edited one-by-one to clean up dead genes, very tedious
• Keep data in flat files? Who will manage the flat file? Xiaodong
• Juancarlos has script to screen all objects for dead genes
• Still need to upload the data (from OA and flat file) for every release/upload/build
• Old interaction objects still in Citace Minus (never removed); Wen will erase
• Keep dead genes in large-scale interactions for now, script their removal later
• Suppress entire objects from loading to ACEDB if has dead genes? No, still dump, but add cautionary remark?
• RNAi interaction objects likely have dead gene assigments (need to check)
• 'Dead genes' 29128, totally 'dead' vs. 'merged' genes?
• Juancarlos will generate list of 'killed' genes (not 'merged')
• Move 'killed' genes to flat files?
• OA will throw error in interaction has only one gene if dumped
• Need to reconcile the fact that the Interaction objects require a GeneID for Interactor tag
• Need to deal with RNAi-to-gene mapping for interactions

Preparation for Arun's departure

• Arun will check all code into SVM repository
• Repository has taken in a bunch of unnecessary data (PDF files, etc.); will avoid from now on
• Daniela will meet with Arun and Yuling, to extract images from PDFs, and marking positive/negative figure legends
• SGD QCFast editor finished (although may not be able to test extensively before Arun leaves)
• Arun will discuss maintenance of QCFast with Michael
• Arun will prepare documentation
• How much work will it take to generate FlyBase editor?
  • Providing capacity/functionality to delete as well as add links in paper

IAIC (International Arabidopsis Information Consortium)

• Imagine 10+ years from now; what data does Arabidopsis researcher have access to?
• People listed features wish-list
• What specifications required to achieve wish-list? APIs, user documentation, code, etc.
• Ideas fall to the side based on difficulties
• Optimistic vs pessimistic approach
• People then start writing specifications in grant format
• At end, 38 page document capturing all ideas/thoughts
• Divide into 4 groups, to pick up all ideas that fell through the cracks
• NSF, MRC, German, Austrian groups
• Community annotation focus
• Core references developed upon consistent, repeated user-generated annotations
• Wikipedia-type monitoring and annotation (minimal staff)
• Moving from single gene queries to multi-gene/process queries
• Looking for appropriate PIs to take charge
• Publicly available document (Jan 2012?)
• NSF has spent $1B on Arabidopsis research, cannot let data just wither

BioCreative group conference

• BioCreative is text mining research/development group
• After BioCurator meeting in April
• Requires submission of 2-page document describing curation workflow
• Kimberly has written a draft and would like feedback (will send out)
• Track II section of BioCreative
• Document = Description of biocurative workflow (for a text miner audience)

Communicating Script function and maintenance pipeline

• How to best document script generation, development, updating, and maintenance

Ian Hope's New Expression data set

• Daniela handling importing of data
• Ian will update us when data is available
• Method paper is out