Clustering Gene Expression Data by Shape Similarity
Author Information
Author(s): T. J. Hestilow, Yufei Huang
Primary Institution: The University of Texas at San Antonio
Hypothesis
Can gene expression data be clustered more effectively by analyzing the shape of expression profiles rather than their magnitude?
Conclusion
The study demonstrates that clustering based on shape similarity can improve the identification of gene clusters compared to traditional methods.
Supporting Evidence
- The proposed method outperformed traditional clustering techniques in identifying gene clusters.
- Clustering based on shape similarity was shown to be more effective than clustering based on expression magnitude.
Takeaway
This study shows that we can group genes by how their expression patterns look over time, which can help us understand their functions better.
Methodology
The study used a Variational Bayes Expectation Maximization algorithm to cluster gene expression data based on the shape of their time-series profiles.
Limitations
The method may not perform well with noisy data and requires careful selection of the number of clusters.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website