Evaluating Methods for Microarray Data Normalization
Author Information
Author(s): André Fujita, João Ricardo Sato, Leonardo de Oliveira Rodrigues, Carlos Eduardo Ferreira, Mari Cleide Sogayar
Primary Institution: University of São Paulo
Hypothesis
Which normalization method is most effective for microarray data?
Conclusion
Support Vector Regression is the most robust method for microarray normalization.
Supporting Evidence
- Support Vector Regression showed the lowest mean square error across different conditions.
- Kernel Regression performed poorly, especially in the presence of outliers.
- Loess, Splines, and Wavelets methods showed similar performance in most cases.
Takeaway
This study looked at different ways to clean up data from gene tests, and found that one method works best at handling mistakes.
Methodology
The study compared five normalization methods using simulated and actual microarray data.
Potential Biases
The Kernel Regression method was found to be highly sensitive to outliers.
Limitations
The study primarily focused on specific types of microarray data and may not generalize to all datasets.
Statistical Information
P-Value
p<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