Project

General

Profile

2011-10-13 conference call

Questions to address

  • start modifying VegBank?
    • make copy because don't have sudo access?
  • first changes to make
  • prototypes for 10/24 working group
  • VegX or simplified VegBank as import schema?: VegX and DwC
  • write VegX to VegBank XSLT?
  • outputs of 10/24 working group
  • other databanks important to integrate with
  • add foreign keys to BIEN 2 database in order to auto-generate ERD?: we have a VegBIEN ERD instead

To do

  • time component
  • good data entry tool
  • login for SALVIAS
  • raw data and actual analysis input data from Brad and John Donahue
  • clear use cases
  • pros and cons of VegBank schema: we're using a modified version of it
  • add specimens and traits capability to VegBank
  • one row per observation
  • taxon traits table

Notes

  • benchmarking/profiling of queries
  • BIEN plone site, iPlant framework site, Redmine issue tracking site
  • scientists need access to Redmine
  • data posted on wiki
  • Brad has main thoughts about status and what to look at
  • reviewing Plone site, VegX schema
  • Bob Peet, Mike Lee: getting access to VegBank
  • for working group meeting: what currently exists, where to go
  • talk to BIEN PIs (e.g. Steve Dolins)
  • useful to have examples of raw data coming in to framework
  • Brad collating raw data
  • processing of data before import into BIEN 2
  • representative materialized result sets that would be useful to scientists
  • things in BIEN useful to have
  • significant overlap of result sets
    • significant different only in # of materialized records
  • need data that you might get from a BIEN query and feed into analyses
    • types of raw data being downloaded from a database
  • data being collected in many different ways
  • import raw data into BIEN with minimal loss of info
  • Bradley Univ efforts: ?
  • need time component in BIEN 2 db
  • current BIEN 2 db allows longitudinal analyses
  • BIEN 2 missing a lot of metadata
  • look at VegBank, VegX
  • don't retain much of BIEN 2 db structure
  • CTFS model better
  • user dataset protection
  • SALVIAS fairly basic; we need something all-encompassing
  • need correct underlying model for scientific info