Global Validation of Microarray Experiments
Author Information
Author(s): Miron Mathieu, Woody Owen Z, Marcil Alexandre, Murie Carl, Sladek Robert, Nadon Robert
Primary Institution: McGill University and Genome Quebec Innovation Centre
Hypothesis
Can a global validation approach improve the assessment of microarray experiments?
Conclusion
The study recommends using random-stratified sampling for validating microarray experiments to enhance accuracy and precision.
Supporting Evidence
- Random-stratified sampling provided more accurate estimates than top-ranked sampling.
- Concordance correlation coefficient (CCC) was recommended as a better validation index.
- Validation results showed that selecting only the largest effects led to underestimating agreement.
Takeaway
This study shows that picking genes randomly for testing is better than just picking the most important ones, making sure we get a clearer picture of how well the tests work.
Methodology
The study used simulated and empirical data to compare different gene selection strategies for validation.
Potential Biases
Non-random sampling strategies may lead to underestimating the agreement between microarray and validation results.
Limitations
The study may not account for all biases in gene selection and validation processes.
Participant Demographics
Mouse NIH 3T3-L1 preadipocytes were used in the experiments.
Statistical Information
P-Value
p<0.005
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website