Clustering Time-Series Gene Expressions Using Spectral Preprocessing
Author Information
Author(s): Wentao Zhao, Erchin Serpedin, Edward R Dougherty
Primary Institution: Texas A&M University
Hypothesis
Can a novel clustering preprocessing strategy that combines spectral estimation techniques with clustering improve the grouping of time-series gene expressions?
Conclusion
The proposed clustering strategy yields significantly different clusters compared to traditional methods and is particularly effective for grouping genes involved in time-regulated processes.
Supporting Evidence
- The proposed method outperformed traditional clustering techniques in grouping cell-cycle genes.
- Simulation results showed that spectral density-based clustering methods achieved better performance than expression-based methods.
- The study validated the clustering schemes by comparing them against known sets of cell-cycle genes.
Takeaway
This study shows a new way to group genes based on their activity over time, which helps scientists understand how genes work together during important processes like the cell cycle.
Methodology
The study used Lomb-Scargle periodogram for spectral density estimation and compared clustering results from hierarchical, K-means, and self-organizing map methods.
Limitations
The study is limited by the small sample size of time-series data and the potential for local optima in clustering algorithms.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website