Clustering cancer gene expression data by projective clustering ensemble.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Gene expression data analysis is crucial for cancer diagnosis and gene treatments.
• Clustering techniques are vital for analyzing high-dimensional gene expression data.
• Challenges include the curse of dimensionality and limited samples in gene expression datasets.

Purpose of the Study:

• To develop a novel method integrating projective and ensemble clustering for gene expression data.
• To address the limitations of existing clustering techniques in handling high-dimensional gene expression data.
• To improve the accuracy and performance of gene expression data clustering, particularly for cancer-related studies.

Main Methods:

• A projective clustering ensemble (PCE) approach was developed.
• PCE integrates projective clustering and ensemble clustering strategies.
• The method was evaluated using publicly available cancer gene expression data.

Main Results:

• PCE demonstrated improved clustering quality by at least 4.5% on average compared to related methods.
• The proposed technique effectively mitigates the curse of dimensionality in gene expression data.
• Experimental results confirmed the superiority of PCE over dimensionality reduction-based single clustering and ensemble approaches.

Conclusions:

• Synergizing projective clustering with ensemble clustering is a promising strategy for enhancing gene expression data analysis.
• Projective clustering ensemble (PCE) offers an effective alternative for clustering cancer gene expression data.
• The findings highlight the importance of combining advanced clustering techniques for complex biological data.