Improving Cross-Platform Correlation in Microarray Studies
Author Information
Author(s): Archer Kellie J, Dumur Catherine I, Taylor G Scott, Chaplin Michael D, Guiseppi-Elie Anthony, Grant Geraldine, Ferreira-Gonzalez Andrea, Garrett Carleton T
Primary Institution: Virginia Commonwealth University
Hypothesis
Can a correction factor improve the estimation of cross-platform correlations in microarray data?
Conclusion
Correcting for measurement error significantly enhances the accuracy of cross-platform correlation estimates in microarray studies.
Supporting Evidence
- The study demonstrated that correcting for measurement error improves cross-platform correlations.
- Intra-platform reproducibility was thoroughly evaluated before estimating cross-platform correlations.
- The Affymetrix platform showed the highest reproducibility among the tested platforms.
Takeaway
This study shows that when scientists compare results from different gene expression tests, they need to fix errors in the measurements to get better results.
Methodology
Microarray hybridizations were performed using three different technologies, and intra- and cross-platform reproducibility was assessed using Pearson's correlation and a correction factor for measurement error.
Potential Biases
The platform-specific sequence of reactions and technician experience may introduce bias in the results.
Limitations
The study's findings are confounded by differences in laboratory protocols and technician experience.
Statistical Information
P-Value
< 0.0001
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website