Assessing Technical Noise in Microarray Data
Author Information
Author(s): Klebanov Lev, Yakovlev Andrei
Hypothesis
How high is the level of technical noise in microarray data?
Conclusion
The analysis suggests that the random fluctuations of gene expression signals caused by technical noise are quite low and have a negligible effect on statistical inference from microarray data.
Supporting Evidence
- The study found that the standard deviations of expression levels were uniformly small across genes.
- The analysis indicated that the observed noise level does not bias correlation coefficients significantly.
- The results challenge the common belief that technical noise is a major issue in microarray data.
Takeaway
This study found that the noise in microarray data isn't as bad as people think, which means the results from these tests are more reliable.
Methodology
The study reanalyzed data from the MicroArray Quality Control (MAQC) Consortium to assess measurement errors in high-density oligonucleotide array technology.
Potential Biases
The study acknowledges that the results may not generalize to other microarray platforms or real-world scenarios where noise levels could be higher.
Limitations
The sample size for each test site was small, and the analysis was limited to the Affymetrix GeneChip platform.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website