Welcome to PhenoFlow

Phenoflow is the ​MOLGENIS reference implementation of the phenotype object model, Pheno-OM, to capture 'any' phenotype observation lead by ​EU-GEN2PHEN. Our mission mission: store individual and summary level observations in a uniform way to enable harmonization and interoperability of phenotypic data across human genetics, model organisms and biobanks. Phenoflow includes an object model, user interface, search wizard, ​ontology links and Excel import/export.

Developers

Core of Phenflow was developed by:

• ​UMC Groningen/GCC (Swertz),
• ​EBI Hinxton/FG (Adamusiak, Parkinson),
• ​FIMM Helsinki (Muilu) and
• ​U Leicester (Thorisson, Brookes)
• ​other members of EU-GEN2PHEN.

Applications using Phenoflow are developed by:

• ​NBIC Biobanking for BiobankCatalog and SemanticSearch? (Antonakaki, van Enckevort, Marshall, Swertz),
• ​LifeLines for BiobankCatalog (Antonakaki, Lops, Swertz)
• ​BioSHARE for biobank data integration (Chao, Swertz)
• ​BBMRI-NL for BiobankCatalog and NgsWorkbench (Antonakaki, Swertz)
• ​U Groningen for ​AnimalDB (Roos, Boerema, Swertz)

Motivation

The next stage of epidemiological and genetic research will depend critically on large collections of high quality samples and data. A wealth of panels from natural human and experimental model organism populations is readily available, but their annotation is scattered between thousands broad inter-institute panel biobanks and deep departmental boutique disease biobanks, each with their own data front-end. While standardization efforts have produced simple formats for the exchange and integration of high throughput data, such as ​MAGE-TAB for microarrays and ​XGAP for xQTL studies an equally lightweight system for phenotypic data is still left wanted

Easy to adopt system

We here report an easy-to-adopt system for the harmonization, integration and search of phenotypic information from experimental and clinical biobanks. The first component is a community created flexible, ontology-enabled model for the uniform representation of any phenotypic data: panels/cohorts, individuals, protocols, observable features and values (Pheno-OM). Then we present a reusable implementation of this model, including a simple tab/excel based data file format (Pheno-TAB) to harmonize, load and share (meta)data on and a ‘database-in-a-box’ (Pheno-DB) for local biobank projects to quickly setup their own phenotype repository or customize it using the ​MOLGENIS system. As a next step we aim to demonstrate how use of ontologies and semantic methods in this system enable integrated search of vast collections of phenotypic information across investigations, species and even a federated network of PhenoDB installations. Thus we expect PhenoFlow to lower the barriers for phenotypic data flow between biobanks and enables integrated search across panels and species to find larger sample sets and data for the next generation of biomedical studies.

A growing range of applications

Currently the PhenoFlow model has been picked up to produce:

• the reference implementation which is available at ​http://www.ebi.ac.uk/microarray-srv/pheno (Adamusiak, Parkinson, GEN2PHEN, Swertz)
• the BiobankCatalog by the ​NBIC Biobanking Task force (Antonakaki, Enckevoort), ​LifeLines (Lops) and ​BBMRI-NL (Swertz).
• the AnimalDb? database for tracking animal observations (Roos)
• the NGS workbench for tracking all around next generation sequencing experiment by ​BBMRI-NL (Lundberg)

The most recent sourcecode of all these projects is available from SVN at ​http://www.molgenis.org/svn/pheno The most recent sourcecode of all these projects is available from SVN at ​http://www.molgenis.org/svn/pheno