Analyzing Tests for Identifying Genes in Lung Cancer
Author Information
Author(s): Jordan Rick, Patel Satish, Hu Hai, Lyons-Weiler James
Primary Institution: University of Pittsburgh
Hypothesis
Can Efficiency Analysis improve the identification of differentially expressed genes in lung adenocarcinoma?
Conclusion
The D1 test was found to be the most consistent and effective method for identifying differentially expressed genes in lung adenocarcinoma.
Supporting Evidence
- The D1 test showed the highest sensitivity and specificity.
- Efficiency Analysis correctly predicted the best test and normalization method.
- Normalization methods did not significantly alter the gene lists unless a p-value criterion was involved.
Takeaway
This study looked at different ways to find important genes in lung cancer and found that one method, called D1, worked the best.
Methodology
The study used publicly available lung adenocarcinoma datasets and applied Efficiency Analysis to compare different normalization methods and tests for identifying differentially expressed genes.
Potential Biases
There is a potential for intensity-related bias in some tests.
Limitations
The study did not consider combinations of normalization methods or other classifiers beyond Naïve Bayes.
Participant Demographics
The Beer dataset included 69 neoplastic lung adenocarcinoma samples and 17 non-neoplastic samples.
Statistical Information
P-Value
0.000391
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website