Related Experiment Videos
Estimating gene networks from gene expression data by combining Bayesian network model with promoter element
Yoshinori Tamada1, SunYong Kim, Hideo Bannai
1Human Genome Center, Institute of Medical Science, The University of Tokyo, Shirokanedai, Minato-ku, Tokyo, Japan.
Bioinformatics (Oxford, England)
|October 10, 2003
Summary
This study introduces a novel statistical method to simultaneously estimate gene networks and detect promoter elements by integrating gene expression and DNA sequence data. The approach improves network accuracy by leveraging transcription factor binding site information.
Area of Science:
- Computational Biology
- Systems Biology
- Bioinformatics
Background:
- Estimating gene regulatory networks from limited microarray data is challenging due to the high dimensionality of the data.
- Accurate gene network inference is crucial for understanding cellular mechanisms.
- Identifying cis-regulatory elements, such as promoter motifs, can provide insights into gene regulation.
Purpose of the Study:
- To develop a statistical method for simultaneous estimation of gene networks and detection of promoter elements.
- To overcome the limitations of estimating gene networks from sparse gene expression data.
- To integrate DNA sequence information into a Bayesian network model for improved network inference.
Main Methods:
- Developed a Bayesian network model integrating microarray gene expression data and DNA sequence information.
- Proposed an iterative approach where network structure is used to detect consensus motifs in promoter regions.
- Re-estimated the gene network using detected motif information until convergence, enhancing accuracy.
Main Results:
- The integrated method demonstrated improved accuracy in gene network estimation compared to methods using expression data alone.
- Successfully identified consensus motifs in promoter regions, linking network structure to regulatory elements.
- Validated the method through Monte Carlo simulations and application to Saccharomyces cerevisiae gene expression data.
Conclusions:
- The proposed statistical method effectively integrates gene expression and DNA sequence data for robust gene network inference.
- Simultaneous detection of promoter elements and network estimation enhances biological interpretability and accuracy.
- This approach offers a powerful tool for dissecting gene regulatory mechanisms in biological systems.