Related Experiment Videos
CUBIC: identification of regulatory binding sites through data clustering
Victor Olman1, Dong Xu, Ying Xu
1Protein Informatics Group, Life Sciences Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6480, USA.
Journal of Bioinformatics and Computational Biology
|August 4, 2004
Summary
We developed a novel computational method to identify transcription factor binding sites by treating the problem as cluster identification in noisy data. Our algorithm efficiently finds these sites and assesses their statistical significance.
Area of Science:
- Bioinformatics
- Computational Biology
- Genomics
Background:
- Transcription factor binding sites regulate gene transcription.
- Computational identification of these sites is challenging.
- Existing methods struggle with noisy biological data.
Purpose of the Study:
- To develop a novel algorithm for identifying transcription factor binding sites.
- To address the challenge of finding clusters in noisy data.
- To provide a statistically robust method for binding site identification.
Main Methods:
- Formulated binding site identification as a cluster identification problem.
- Utilized sequence similarity and frequency of binding sites.
- Developed a novel algorithm for cluster identification in noisy backgrounds.
- Implemented a statistical significance assessment method.
Main Results:
- Presented a novel algorithm for identifying clusters in noisy data.
- Demonstrated efficient and rigorous solving of cluster identification problems.
- Developed a method to assess statistical significance and rule out accidental clusters.
- Implemented the algorithm and methods into computer software CUBIC.
Conclusions:
- The CUBIC software provides an effective solution for identifying transcription factor binding sites.
- The novel algorithm successfully addresses the challenge of cluster identification in noisy data.
- The statistical significance assessment enhances the reliability of identified binding sites.