Functional clustering of time series gene expression data by Granger causality

Area of Science:

• Bioinformatics
• Systems Biology
• Computational Biology

Background:

• Traditional gene expression analysis often clusters genes solely based on similar temporal patterns.
• Genes function within complex networks, suggesting that expression similarity alone may not fully capture functional relationships.
• Identifying functionally similar genes is crucial for understanding cellular processes.

Purpose of the Study:

• To explore the utility of Granger causality for clustering time series gene expression data.
• To identify functionally related genes by incorporating network topological features.
• To propose a complementary approach to traditional gene clustering methods.

Main Methods:

• Gene clustering was performed using Granger causality analysis on time series gene expression data.
• Granger causality identifies directed relationships between time series, indicating potential causal influence.
• Analysis considered both between-set and within-set Granger causality.

Main Results:

• The study demonstrates the application of Granger causality for gene clustering.
• Identified causal relationships provide insights into gene interactions beyond mere expression correlation.
• This method allows for the construction of gene networks based on inferred causality.

Conclusions:

• Granger causality offers a novel method for functional gene clustering.
• Clustering based on Granger causality reveals topological proximity and functional similarity.
• This approach complements expression-based clustering, providing a more comprehensive understanding of gene function in biological networks.