Unsupervised clustering in mRNA expression profiles.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Microarray technology enables high-throughput analysis of gene expression.
• Analyzing large-scale microarray data presents significant computational challenges.
• Cluster analysis is a common approach for gene expression data, but determining the optimal number of clusters can be difficult.

Purpose of the Study:

• To investigate the efficacy of the k-windows clustering algorithm for gene expression microarray data analysis.
• To evaluate the algorithm's ability to automatically determine the number of clusters.
• To enhance clustering quality using dimension reduction techniques.

Main Methods:

• Application of the k-windows clustering algorithm to gene expression data.
• Utilizing various dimension reduction techniques to preprocess the data.
• Proposing and implementing a hybrid dimension reduction approach.

Main Results:

• The k-windows algorithm successfully identified clusters within the gene expression data.
• The algorithm automatically determined the optimal number of clusters without prior knowledge.
• The integration of dimension reduction techniques improved the overall clustering performance.
• High classification success rates were achieved.

Conclusions:

• The k-windows clustering algorithm is a viable and effective tool for analyzing gene expression microarray data.
• The algorithm's ability to determine cluster numbers simplifies the analysis process.
• Hybrid dimension reduction strategies can significantly enhance the accuracy of gene expression clustering.