iRefIndex

Overview

iRefIndex was a consolidated protein interaction database that provided a unified index for protein interaction data from multiple primary databases. The resource used sequence-based hash keys (computed using the Secure Hash Algorithm) to uniquely identify protein interaction records and group together redundant data, while carefully tracking the provenance of all integrated information.

Note: The iRefIndex website appears to be inactive as of 2025. However, the methodology and historical data remain valuable references for the field, and the consolidation approach influenced subsequent protein interaction data integration efforts.

Key Innovation

iRefIndex introduced a rigorous method for:

1. Generating unique keys for protein interaction records based on protein primary sequences and taxonomy identifiers
2. Identifying redundant groups where multiple records describe the same molecular interaction
3. Recording provenance by tracking all consolidation operations with mapping scores
4. Maintaining quality feedback by reporting problematic references (malformed, deprecated, ambiguous, or unfound) back to source databases

Integrated Databases

iRefIndex consolidated data from multiple primary interaction databases:

• BIND (Biomolecular Interaction Network Database)
• BioGrid
• DIP (Database of Interacting Proteins)
• HPRD (Human Protein Reference Database)
• IntAct
• MINT (Molecular INTeraction database)
• MPact
• MPPI
• OPHID

Data Format

The consolidated interaction reference index was provided in PSI-MITAB 2.5 format, a standardized tabular format for molecular interactions developed by the HUPO Proteomics Standards Initiative.

Associated Tools

• iRefR: R package for programmatic access and manipulation of iRefIndex data
• iRefScape: Cytoscape plugin for visualization and data mining
• iRefWeb: Web-based interface for navigating the protein-protein interaction landscape

Use Cases

• Meta-analysis of protein interactions across multiple data sources
• Removing redundancy when combining interaction datasets
• Training and validation of interaction prediction methods
• Network analysis requiring comprehensive interaction data
• Referenced by major resources like GeneMANIA, Reactome, and STRING

Historical Impact

iRefIndex was widely adopted by the bioinformatics community and influenced major resources including:

• GeneMANIA (gene function prediction)
• Reactome (pathway database)
• STRING (protein association networks)
• Various computational biology tools and methods

Publications

The iRefIndex methodology and applications were described in multiple peer-reviewed publications:

1. Main paper: Razick S, Magklaras G, Donaldson IM. “iRefIndex: a consolidated protein interaction database with provenance.” BMC Bioinformatics. 2008;9:405. PMID:18823568

2. R package: Mora A, Donaldson IM. “iRefR: an R package to manipulate the iRefIndex consolidated protein interaction database.” BMC Bioinformatics. 2011;12:455. PMID:22115179

3. Cytoscape plugin: Razick S, Mora A, Michalickova K, Boddie P, Donaldson IM. “iRefScape. A Cytoscape plug-in for visualization and data mining of protein interaction data from iRefIndex.” BMC Bioinformatics. 2011;12:388. PMID:21975162