Not proper ROC curves as new tool for the analysis of differentially expressed genes in microarray experiments
2008

New Method for Analyzing Gene Expression in Microarray Experiments

Sample size: 34 publication Evidence: moderate

Author Information

Author(s): Stefano Parodi, Vito Pistoia, Marco Muselli

Primary Institution: G. Gaslini Children's Hospital, Genoa, Italy

Hypothesis

Can a new method based on the area between the ROC curve and the rising diagonal (ABCR) identify differentially expressed genes that standard methods miss?

Conclusion

The new method identified 1607 differentially expressed genes, including 16 that corresponded to not proper ROC curves, which standard methods failed to detect.

Supporting Evidence

  • The new method identified 1607 differentially expressed genes with a 15% estimated False Discovery Rate.
  • Among the identified genes, 16 corresponded to not proper ROC curves that standard methods missed.
  • The method successfully distinguished between subclasses of normal B cells and heterogeneous lymphomas.

Takeaway

Researchers created a new way to find important genes in experiments that look at how genes behave under different conditions, and it found many genes that other methods missed.

Methodology

The study used a new statistical method combining standard ROC analysis with a new approach based on ABCR and TNRC to identify differentially expressed genes.

Potential Biases

The study may have biases related to the selection of genes based on statistical thresholds.

Limitations

The TNRC method may have low statistical power in small sample sizes, potentially missing some differentially expressed genes.

Participant Demographics

The study included 14 normal B cell samples and 20 heterogeneous lymphoma samples.

Statistical Information

P-Value

0.05

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1471-2105-9-410

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication