Functional regression clustering with multiple functional gene expressions

Area of Science:

• Bioinformatics
• Computational Biology
• Systems Biology

Background:

• Gene expression data analysis often involves time-series experiments.
• Identifying genes with similar expression pattern responses across conditions is crucial but challenging.

Purpose of the Study:

• To develop a novel method for clustering genes with similar temporal expression pattern relationships, even with differing individual profiles.
• To apply this method to analyze diurnal gene expression patterns perturbed by seasonal changes.

Main Methods:

• A K-means-type algorithm utilizing function-on-function regression models.
• The model accommodates multiple functional explanatory variables for robust clustering.
• Extensive simulations were performed for validation.

Main Results:

• The proposed method successfully identified clusters of genes with similar expression pattern relationships.
• Application to diurnal gene expression revealed seasonal perturbations.
• Identified clusters were enriched for genes with related biological functions, including photosynthesis and polysomal ribosomes.

Conclusions:

• The novel clustering approach provides useful and novel biological insights.
• It effectively groups genes based on shared response patterns, not just individual expression levels.
• The method has potential applications in various fields of gene expression analysis.