Integrating Microarray Data for Better Analysis
Author Information
Author(s): Yinglei Lai, Sarah E Eckenrode, Jin-Xiong She
Primary Institution: The George Washington University
Hypothesis
Can a statistical method improve the integration of different microarray data sets for differential expression analysis?
Conclusion
The study shows that not considering genome-wide concordance can lead to misleading results, and the proposed method is robust and useful for integrative analysis.
Supporting Evidence
- The method provides a rigorous parametric solution for data integration.
- Simulation studies confirm the importance of evaluating genome-wide concordance.
- The proposed method outperforms traditional pooling approaches when data sets are not concordant.
Takeaway
This study helps scientists combine data from different experiments to find important genes, but they need to make sure the data is similar first.
Methodology
The study uses a mixture-model based statistical framework to evaluate genome-wide concordance and perform data integration.
Potential Biases
Potential biases may arise from the covariance structure of microarray data.
Limitations
The method assumes independence among test scores and may not validate the assumed mixture model.
Participant Demographics
The study involved data from NOD mice and prostate cancer subjects.
Statistical Information
P-Value
p<0.01
Statistical Significance
p<0.01
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website