Random walk models for bayesian clustering of gene expression profiles.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Analyzing gene expression temporal profiles is crucial in functional genomics.
• Model-based clustering captures data dynamics and identifies optimal clusters.
• Existing methods struggle with time dislocations, non-stationarity, and irregular temporal grids.

Purpose of the Study:

• To develop a novel Bayesian method for analyzing gene expression temporal profiles.
• To address limitations of current approaches, including time dislocations, non-stationarity, and irregular data.
• To provide a robust tool for functional genomics research.

Main Methods:

• A new Bayesian method based on random walk models was developed.
• The method explicitly models inter-gene variability within clusters.
• Mild a priori assumptions are required regarding data-generating processes.

Main Results:

• The method was validated on simulated datasets.
• It was successfully applied to analyze serum-stimulated fibroblast gene expression data.
• Promising results indicate the method's utility in functional genomics.

Conclusions:

• The developed Bayesian method effectively handles complex gene expression temporal profiles.
• It addresses key challenges like time dislocations and irregular sampling.
• This approach shows significant potential to advance functional genomics research.