Jamboree2012

Projects

Project 1: dbXP developers getting started

Philip, Chris, Jeroen, Guido, Bernhard

1a Make it easier to get started with deploying GSCF and the modules

• Creation of an Ubuntu virtual machine: done by Guido. Download will be made available by Jeroen.
• Description of deployment environment: see http://phenotypefoundation.github.com/GSCF
• Bernhard managed to install GSCF on CentOS with some modifications to the install guide, which will be incorporated. Also, running GSCF without internet access (company firewalls) is a problem, so we need to implement proxy functionality.

1b Create a Hello World module for dbXP

• How to connect your own processing pipeline to GSCF (how to connect your pipeline to the Phenotype Database) ** architecture of the system ** documentation of API's ** create a Hello World module Is started but not finished, see https://github.com/keesvanbochove/dbxpHelloWorld

Project 2: dbXP bioinformaticians / users getting started

Thomas, Ferry, Anwesha, Lars

2a Make it easier to get started with data from GSCF+modules in R

Finished, see https://github.com/PhenotypeFoundation/GSCF-RClient

2b Document how to leverage PathVisio RPC to analyze dbXP data

Ferry and Anwesha ended up integration PathVisio functionality into GSCF and gave a working demonstration. Ferry will create a new branch in GitHub with the code, which will be integrated in due time as pathway visualization functionality.

2c Create a Bioclipse plugin

Was started by Andra, see https://github.com/PhenotypeFoundation/dbXP_BioclipsePlugin

Project 3 Create an ISATAB exporter in GSCF - development project

Andra, Kees -- ISATAB --> how to encode the study design in ISATAB --> create an exporter to ISATAB, and document what we had to discard or flatten in the process Was started by Andra by creating a Bioclipse plugin leveraging the GSCF API. Stuck with Java libraries for REST/HTTP/JSON, so we might want to switch to Groovy for that.

Additional comments

• RDF --> too big to take on now, but for the roadmap: a first step would be to export the data as-is in RDF, encoding the relations in accordance with the current data model (Andra)
• Integration with Synapse could be a good match, Synapse provides a rich platform centered around analyses while Phenotype Database mainly revolves around the data
• It would be useful to create a link from analyses to the study data that you used (and which is in GSCF + modules)