AMICa

AMICa performs multi-method clustering and analysis of microarray gene expression data to identify co-expressed genes, suggest optimal cluster numbers, and assess clustering quality.

Key Features:

• Multi-Method Clustering Engine: Implements a variety of clustering algorithms and allows concurrent application and comparison of multiple methods on the same dataset.
• Automatic File Format Detection: Detects input file formats automatically to streamline data import for microarray gene expression datasets.
• Cluster Number Suggestions: Provides suggestions for the optimal number of clusters using a variant of the stability-based method proposed by Tibshirani et al.
• Data Visualization: Produces heatmaps for visual inspection of gene expression patterns and cluster structures.
• Clustering Quality Assessment: Reports clustering quality measures such as cluster homogeneity.
• Efficient Large-Scale Processing: Supports algorithms optimized for large datasets, including FPF-SB and k-Boost, and a batch-mode to run selected algorithms simultaneously.

Scientific Applications:

• Gene Expression Profiling: Identification of co-expressed genes from microarray datasets.
• Disease Classification and Biomarker Discovery: Classification of diseases based on gene expression patterns and support for biomarker identification.
• Functional Genomics Studies: Exploration of functional relationships between genes through cluster analysis.

Methodology:

Uses multiple clustering algorithms, a variant of the Tibshirani et al. stability-based method for cluster number selection, automatic file format detection, heatmap visualization, clustering quality metrics (e.g., cluster homogeneity), and algorithms optimized for large datasets such as FPF-SB and k-Boost, with an option to run selected algorithms in batch mode.