Related Experiment Videos
Clusterv: a tool for assessing the reliability of clusters discovered in DNA microarray data.
1DSI, Dipartimento di Scienze dell'Informazione, Università degli Studi di Milano, Via Comelico 39, Italy. valentini@dsi.unimi.it
Bioinformatics (Oxford, England)
|December 8, 2005
Summary
We developed an R package to evaluate the reliability of clusters found in high-dimensional DNA microarray data. It uses random projections to maintain data distances, improving cluster analysis accuracy.
Area of Science:
- Bioinformatics
- Computational Biology
- Genomics
Background:
- High-dimensional DNA microarray data presents challenges for cluster analysis.
- Assessing the reliability of discovered clusters is crucial for biological interpretation.
Purpose of the Study:
- To introduce a novel R package for reliable cluster assessment in high-dimensional DNA microarray data.
- To provide a computational tool for validating clustering results.
Main Methods:
- Implementation of methods based on random projections.
- Approximation of distance preservation in projected subspaces.
Main Results:
- The R package enables robust assessment of cluster reliability.
- Random projection methods effectively maintain inter-example distances.
Conclusions:
- The package offers a valuable resource for analyzing DNA microarray data.
- The employed random projection techniques enhance the reliability of cluster discovery.