From microarray to biological networks: Analysis of gene expression profiles.

Area of Science:

• Systems Biology
• Computational Biology
• Genomics

Background:

• Gene expression profiling generates vast datasets from various cellular challenges.
• Dissecting complex genetic networks controlling gene expression is crucial.
• Analyzing time-series gene expression data requires robust computational methods.

Purpose of the Study:

• To present a general approach for developing dynamic models to analyze time-series whole-genome expression data.
• To enable the dissection of complex genetic networks.
• To visualize gene regulatory interactions.

Main Methods:

• Developing dynamic models for time-series gene expression analysis.
• Utilizing singular value decomposition (SVD) to calculate model parameters, representing gene influence.
• Generating correlative networks using a threshold approach.
• Integrating dynamic models with cluster analysis (two-way hierarchical clustering) for visualization.

Main Results:

• The dynamic model parameters quantify the influence of one gene's expression on another.
• Correlative gene networks were generated based on these parameters.
• Hierarchical clustering effectively visualized how genes influence each other's expression levels.
• The approach was successfully demonstrated on yeast cell cycle gene expression data.

Conclusions:

• Dynamic models provide a powerful framework for analyzing complex gene expression time-series data.
• This method facilitates the identification of gene regulatory relationships.
• The integration with clustering offers insights into dynamic gene interactions within cellular processes.