Related Experiment Videos
springScape: visualisation of microarray and contextual bioinformatic data using spring embedding and an 'information
Timothy M D Ebbels1, Bernard F Buxton, David T Jones
1Bioinformatics Unit, Department of Computer Science, University College London, Gower Street, London, WC1E 6BT.
Bioinformatics (Oxford, England)
|July 29, 2006
Summary
springScape integrates high-throughput data with biological context for enhanced interpretation. This system visualizes complex relationships in gene expression and protein interactions, aiding discovery.
Area of Science:
- Bioinformatics
- Computational Biology
- Systems Biology
Background:
- Interpreting high-throughput data requires biological context, which standard analysis tools often fail to provide.
- Integrating diverse data sources like microarrays, protein interactions, and gene ontology (GO) annotations is challenging.
Purpose of the Study:
- To develop a novel system, springScape, for integrating and visualizing multiple biological data sources.
- To enable dynamic exploration of complex relationships within high-throughput and bioinformatic data.
- To improve the interpretation of microarray data by incorporating relevant biological context.
Main Methods:
- Developed springScape, a system utilizing 'spring embedding' and an 'information landscape' visualization.
- Represented each data source as a network of nodes and weighted edges.
- Combined networks using spring embedding for similarity-based node proximity.
- Employed modified Procrustes analysis to ensure visualization reproducibility.
- Integrated microarray data with protein-protein interaction data and Gene Ontology (GO) annotations.
Main Results:
- springScape dynamically combines multiple data sources, highlighting specific features.
- Visualizations reveal complex relationships by adjusting data source weights and observing network dynamics.
- Identified spatio-temporal gene expression profiles.
- Discovered GO terms correlated with gene expression and protein interactions.
- Demonstrated acceptable reproducibility of visualizations.
Conclusions:
- springScape offers a promising tool for interpreting microarray data within a broader bioinformatic context.
- The system leverages interactive feedback and human visual pattern recognition for biological discovery.
- Dynamic integration and visualization of diverse data enhance the understanding of complex biological systems.