Difference between revisions of "Hackathon 2013/BestPractices"

From TaxonWorks Wiki
Jump to: navigation, search
 
(One intermediate revision by one user not shown)
Line 1: Line 1:
 +
What we did:
 +
Assessed other taxonomic workbenches, past and present, to try to ID reasons for success/sunsetting, and then glean best practices/things to look out for as a result.
 +
 +
Overall the systems that have been developed seem to fall victim to funding fall-through and shifting of focus and leadership.  The systems that have survived seem to be:
 +
* open source and free
 +
* developed by or in coordination with practicing taxonomists
 +
* have steady funding or support (i.e. independent of other grants or projects)
 +
* are modular and support later changes in design and alteration (Diversity Workbench and Scratchpads in particular)
 +
* are designed to support collaboration between multiple people (this isn't necessarily something that is required for success, though -- however it's interesting that they seem to be designed for use cases that require collaboration at a distance)
 +
* seem to be cloud-based, though am unclear on that for a few of them
 +
* note: couldn’t look at ITIS taxonomic workbench because of gov’t shut down.
 +
 +
Basic best practices and system needs from conversation with Lisa:
 +
* must provide offline access to data (so entirely cloud based is not great)
 +
* support rapid data entry (either through import or clear and easy support of transcription)
 +
* should be able to search from other repositories (GenBank)
 +
* available on one machine (again, cloud based is not great)
 +
* must support publication workflow (thus, good reference management modules are important)
 +
* need to be able to hook into institutional specimen databases (depending on workflow.  This means that a workbench either needs to be a solution for tricky institutional database setups (e.g. idiosyncratic DBS or convoluted KE Emu set ups), or something that can work alongside them)
 +
* for systematists, should fit into existing/common bioinformatics pipelines
 +
 +
 
Here is our working google doc: [https://docs.google.com/document/d/14cJvWjdneQMIcO3RfS_EEDpethQXZcmY5VII1GruFwY/edit?usp=sharing]
 
Here is our working google doc: [https://docs.google.com/document/d/14cJvWjdneQMIcO3RfS_EEDpethQXZcmY5VII1GruFwY/edit?usp=sharing]
 +
And the spreadsheetized version: [https://docs.google.com/spreadsheet/ccc?key=0Agy0Ax3YY340dG9PejNPTkZ2Q2RrcjU1WVZMYWxYT2c&usp=drive_web#gid=0]
  
 
== Members ==
 
== Members ==

Latest revision as of 11:01, 4 October 2013

What we did: Assessed other taxonomic workbenches, past and present, to try to ID reasons for success/sunsetting, and then glean best practices/things to look out for as a result.

Overall the systems that have been developed seem to fall victim to funding fall-through and shifting of focus and leadership. The systems that have survived seem to be:

  • open source and free
  • developed by or in coordination with practicing taxonomists
  • have steady funding or support (i.e. independent of other grants or projects)
  • are modular and support later changes in design and alteration (Diversity Workbench and Scratchpads in particular)
  • are designed to support collaboration between multiple people (this isn't necessarily something that is required for success, though -- however it's interesting that they seem to be designed for use cases that require collaboration at a distance)
  • seem to be cloud-based, though am unclear on that for a few of them
  • note: couldn’t look at ITIS taxonomic workbench because of gov’t shut down.

Basic best practices and system needs from conversation with Lisa:

  • must provide offline access to data (so entirely cloud based is not great)
  • support rapid data entry (either through import or clear and easy support of transcription)
  • should be able to search from other repositories (GenBank)
  • available on one machine (again, cloud based is not great)
  • must support publication workflow (thus, good reference management modules are important)
  • need to be able to hook into institutional specimen databases (depending on workflow. This means that a workbench either needs to be a solution for tricky institutional database setups (e.g. idiosyncratic DBS or convoluted KE Emu set ups), or something that can work alongside them)
  • for systematists, should fit into existing/common bioinformatics pipelines


Here is our working google doc: [1] And the spreadsheetized version: [2]

Members

  • Andrea Thomer
  • Daisie Huang
  • David Eades
  • Julie Allen