Bayesian Detection of Periodic mRNA Time Profiles
Author Information
Author(s): Claes R. Andersson, Anders Isaksson, Mats G. Gustafsson
Primary Institution: Uppsala University
Hypothesis
Can a learning-free Bayesian detector effectively identify periodically expressed genes from microarray data without using training examples?
Conclusion
The Bayesian detector outperforms alternative methods and reveals that only a small fraction of detected genes are cell-cycle related.
Supporting Evidence
- The Bayesian detector showed improved performance over Fisher's g-test and a combination test.
- Only 12% of the detected genes were confirmed as cell-cycle related.
- The study identified 19 new motifs potentially involved in periodic expression.
Takeaway
This study shows a new way to find genes that turn on and off in cycles, which helps us understand how cells grow and divide.
Methodology
The study used a Bayesian detector to analyze simulated and real mRNA expression data from yeast cells during the cell cycle.
Potential Biases
Potential biases may arise from the assumptions made about the period time and the nature of the simulated data.
Limitations
The detector's performance is limited by the lack of labeled training data and the variability of gene expression across different conditions.
Participant Demographics
The study focused on mRNA expression data from Saccharomyces cerevisiae (budding yeast).
Statistical Information
P-Value
0.001
Statistical Significance
p<0.001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website