View
337
Download
0
Embed Size (px)
DESCRIPTION
II Jornades de Consultoria Estadística i Software - Octubre 2013
Citation preview
InformationInformation Management (Management (andand more) at more) at
thethe StatisticsStatistics andand BioinformaticsBioinformatics Unit, Unit,
VallVall dd’’HebrHebróónn InstitutInstitut de Recerca (VHIR) de Recerca (VHIR)
withwith thethe WikiWiki--Tiki CMS Tiki CMS systemsystem
Alex Sánchez1,2, Xavier de Pedro2, Ferran Briansó2
1Departament d’Estadística UB 2Statistics and Bioinformatics Unit (VHIR)
Vall d’Hebron Research Institute
Unitat d’Alta Tecnologia
Statistics and Bioinformatics Unit
4
UEB services
Data analysisOMICS & STandard
Bioinformatics Training Consulting
•Microarrays,
• RTqPCR
•Methylation
•Metagenomics
•Exome sequencing
•Clinical / Epidemiological
• Web & Databasedevelopment
• HPC services
• Pipelines development
• Short Workshops
•Courses
•Official training (MSc)
•Sample size
•Experimental design
•GRANT review
•Statistical parts ofpapers
We provide support in …
UEB resources – We work with ...
UEB goals – We aim at...
• We aim at providing high quality support– Services / Innovation / Teaching
• How can we get to it?– We need highly qualified technicians
– State of the art tools & resources (OS)
– A profesional approach to service• Provide value for money
• Produce well documented, reproducible, results
• Manage projects and processes adequately
– We are working to obtain ISO-9001
UEB Projects & Processes
• Main type of projects
- Omics Data Analysis
- Information management
• Information flow
- Meetings
- Analysis/Reports/Results
- Databases Customers/Studies
- Process management
- Documentation/Protocols
1. Accounting for quality in studies & projects
• Create automatic & reproducible workflows.
– Easy to use by different (expert) technicians
– Possibility to reproduce
• Identically for verification
• With variations for process tuning
– Maintained with Bazar (Version Control System)
• Different technologies
– R / Latex / Sweave(knitr)
– perl / shell programs
– Taverna / Galaxy
2. Controlling information flows
First wave
• Several systems
– Google • Docs, calendars
• Spreadsheets
– Dropbox
– MS-Acces databases
Next wave (Tiki)
• One unique system– Very good for
• Documentation
• File/Image sharing
• Event /Calendar management.
– Trackers for
• databases
• Tasks, projects
– Extra bonus!
• R web applications
The Tiki-Wiki CMS
Main Features
The wiki-way
Features 1 - Wiki pages
• Simple editor– Fonts, Lists, Headings
– Links, Tables. Images
• Markdown-like or wysiwig
• Real time edit (no uploads)
Ex. 1- A simple wiki page
Features 2 – Wiki structures
• Facilities to manage page structures– Add pages easily
– Structure layout
– Table of contents
Example 2- UEB page layout
Features 3 –Granular permissions system
Features 4 – Calendar & Events
Features 5 - Files
Web folders from Tiki file galeries
Web Dav: direct access from office apps
Features 6 - Trackers
Online databases
Form & database tightly integrated
Use it to “track”processes.
Easy to configure views
Examples 6- UEB task trackers
• Other UEB trackers
- Studies,
- Projects
- Customers
- Consultings
Alltogether: ueb.vhir.org (public)
Alltogether: ueb.vhir.org (internal)
In summary
– Using the Tiki-Wiki CMS information flows happening in our unit can be managed.• Wiki pages for
– Documentats & Protocols
– Teaching materials
• Trackers for
– Studies & Customers databases
– Project & Task management
• Calendars & News for planning and informing
• File & Image galleries for safestoring
– All finely customized thanks to permissions
There is more than this
• A common important need is to be able to use R from the web– Internally: Create GUI to automate process
• Call pipelines
• Collect results
– Externally• Provide common-use tools to users
– Increase visibility
– Saves us from repeating monotonous tasks
» E.g. Post exploiting results
R in the web: Many tools, no winner
• Many tools– Recent
• Shiny, Concert, rook, Rapache
– Older• CGIwithR, Rphp,
• Usually hard to implement
Our choice (Tiki + pluginR)
Example: ueb.vhir.org/heatmaps
Example: Basic Pipe
• R/Sweave
• Covers all steps for a MDA
• Start with a (long) parameters file
• Run R/Sweave
• Output– Results
• Tables
• Figures
– Report
– HTML browser
• Use trackers/forms to– Build parameters file
– Control execution/iteration
• Use File/Image galleries to– Store results, plots
• Use wiki page to present– Report
– Give access to results
Current version Web version
In summary (2)
• The Wiki-Tiki CMS has interesting capabilities to empower efficient work at all levels.
• In a context where it can be necessary to create web pages implementing R programs, combining Tiki with R through pluginR is a powerful approach.
Acknowledgments
� The members of the Unitat d’Estadística i Bioinformàtica of the VHIR
https://ueb.vhir.org
� Xavi de Pedro – Tiki developer
� Ferran B.
� J.L Mosquera Tiki testers/users
� Ricardo Gonzalo
� Unitat d’alta tecnologia at VHIR
https://uat.tiki.org
� Grup de Recerca en Estadística i Bioinformàtica (UB)
https://eib.stat.ub.edu