An effective tri-clustering algorithm combining expression data with gene regulation information.

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Bi-clustering identifies genes with similar expression patterns but struggles with direct interpretation of regulatory mechanisms.
• Integrating gene expression data with transcription factor binding information is crucial for understanding gene regulation.
• Existing methods lack a unified approach to combine these data types for robust analysis.

Purpose of the Study:

• To develop a novel computational method for integrating gene expression and gene regulatory information.
• To enhance bi-clustering by incorporating transcription factor binding data for improved biological insight.
• To enable the identification and analysis of dynamic gene regulatory networks.

Main Methods:

• Defined Regulated Expression Values (REV) by combining gene expression and regulatory data.
• Extended bi-clustering to a 3D space using a heuristic TRI-Clustering algorithm.
• Developed the Automatic Boundary Searching (ABS) algorithm to determine cluster boundaries.

Main Results:

• TRI-Clustering and ABS efficiently detect tri-clusters using integrated data.
• Analysis of yeast sporulation data revealed significant gene expression differences in specific stages.
• Reconstructed regulatory networks showed dynamic changes in transcription factor activity during sporulation.

Conclusions:

• The proposed method effectively integrates gene expression and regulatory data for robust clustering.
• The approach facilitates the study of dynamic regulatory mechanisms in biological processes.
• This method offers a generalizable framework for analyzing transforming gene regulatory networks.