WormBase-Caltech Weekly Calls April 2016

From WormBaseWiki
Jump to navigationJump to search

April 7, 2016

Tissue-enrichment tool

  • Juancarlos and Raymond working on
  • How to best integrate this into WormBase?
  • Best to have all production code/material in one place, but could set up a proxy if it's already up and running
  • Considering we don't know yet how popular the tool will be, the proxy could be a sufficient first step
  • Could be a subdomain, e.g. tissueenrichement.wormbase.org, similar to the data submission forms
  • Is this adaptable for the other MODs/organisms? Will submit the paper first

Karen's talk at ZFIN

  • Talk went well
  • some interest with SVM
  • Having good discussions about data types and nomenclature
  • Have their own reagent mapping pipeline to their genome browser

Daniela starting on gene regulation curation

  • Daniela working with Xiaodong to learn the curation process
  • Xiaodong's last day, next Friday 16th

April 21, 2016

Author submission forms

  • Cell Press generating simple 3-column table asking for reagents
  • Likely has a for-profit incentive; we need to push data tables that aid the MODs
  • Chris will send around the table for minimum information of phenotype data
  • Chris will look into a similar table for interactions

FTP file content changes

  • GAFs, for example, get changed
  • We need a mechanism/protocol to make notes of changes to file formats, etc.
  • We could post timestamp and signature entries to a READ_ME file in each FTP directory for all changes to files
  • In addition, we can post on the WB homepage and bog
  • Ranjana will email Todd and Kevin to ask about best practices
  • Ranjana will then send email to staff once a decision has been made


April 28, 2016

joint mod team meeting (ky)

  • ZFIN, SGD, WB: Next Wednesday May 4. All WormBase invited
  • Doug Howe will send around the announcement with links to some google docs
  • SGD will be supplying video conferencing through Blue Jeans http://bluejeans.com/video-conferencing
  • Introductions of each MOD team member and overview of the forum
  • Doug, Karen, and possible SGD person will spend 10-15 minutes introducing the respective MOD website, very basic navigation and search.
  • will invite all mods
  • may be good to record the meeting for sharing and later review

InterMOD (Intermine) call (ky)

  • Getting ready for the TAGC conference in July
  • Creating a single portal for mining across all MODmine instances
  • Current problem is deciding on which homology sets to use, different mods have different source preferences for these homolog sets. e.g. elegans uses Compara, Ortho-MCL, InParanoid, while there other groups may also use these tools, they don't necessarily use all of them and some use different ones, ex. FlyBase uses DIOPT. different homolog sets give different results.

Mailing lists

  • We would like to have a UMOD mailing list
  • Human disease group already has a cross-MOD mailing list
    • Human disease group meeting at Caltech May 16,17

Tissue enrichment analysis (tea) tool

Sentence-level SVM

  • Trying to predict sentences that have phenotype statements for curation
  • Created an automated feature word list by training on ~800 sentences
  • This automated feature list has good but variable success (precision from 48% to 95%, and recall from 80% to 90%)
  • Chris generating a manually generated empirical feature list based on manually reviewed sentences; will assess success rate soon
  • Capitalization of feature words appears to be an issue; would be best if the SVM predictions used case-insensitive searching