Evaluating Microarray Data Normalization Methods
Author Information
Author(s): Patrik Rydén, Henrik Andersson, Mattias Landfors, Linda Näslund, Blanka Hartmanová, Laila Noppa, Anders Sjöstedt
Primary Institution: Umeå University
Hypothesis
Can spike-in experiments effectively compare different microarray data normalization methods?
Conclusion
Spike-in experiments are effective for evaluating microarray preprocessing procedures, revealing a trade-off between sensitivity and bias.
Supporting Evidence
- Spike-in experiments allow for the characterization of analyzed data.
- High sensitivity was observed in analyses using the constrained model or partial filtration.
- Most analyses underestimated the magnitude of gene regulations.
Takeaway
This study looked at different ways to clean up data from gene tests to see which methods work best. They found that some methods are better at finding important genes than others.
Methodology
The study evaluated 252 normalization methods using data from eight spike-in microarrays.
Potential Biases
The methods may underestimate the true regulation of differentially expressed genes, especially at low concentrations.
Limitations
The evaluation is based on a limited number of arrays and does not account for biological variation.
Statistical Information
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website