GEDAS

GEDAS integrates multiple preprocessing methods, distance measures, clustering and classification algorithms to standardize analysis of microarray gene expression data for comparative and exploratory genomics.

Key Features:

• Integration of Multiple Tools: Consolidates numerous algorithms and methodologies within a single suite for consistent analysis of microarray gene expression data.
• Diverse Distance Measures and Pre-processing Techniques: Includes an extensive collection of distance measures and preprocessing techniques for flexible handling of raw microarray data.
• Support for Multiple Algorithms: Supports k-means clustering, hierarchical clustering (HC), singular value decomposition/principal component analysis (SVD/PCA), support vector machines (SVM), Kohonen's self-organizing maps (SOM), and learning vector quantization (LVQ).
• Standardization: Brings together various tools and techniques to reduce variability and enhance reproducibility in microarray data analysis.
• Software Lineage: Implemented as an extension of Cluster 3.0 and builds upon Eisen Lab's original Cluster and Tree View software.

Scientific Applications:

• Pattern discovery: Identifying gene expression patterns in microarray and high-throughput gene expression datasets.
• Sample classification: Classifying samples based on gene expression profiles using clustering and supervised methods such as SVM.
• Exploratory analysis of mechanisms: Exploring underlying biological mechanisms driving observed phenotypes through dimensionality reduction (SVD/PCA) and clustering approaches.

Methodology:

Pre-processing of raw microarray data using selected techniques, application of clustering and classification algorithms (k-means, HC, SOM, LVQ, SVD/PCA, SVM), and comparative analysis using multiple distance measures.