Comparing Methods for Calculating Gene Expression from Microarrays
Author Information
Author(s): Millenaar Frank F, Okyere John, May Sean T, van Zanten Martijn, Voesenek Laurentius ACJ, Peeters Anton JM
Primary Institution: Utrecht University
Hypothesis
The signal of gene expression, and therefore the identity of genes that are significantly up/down regulated, depends on the algorithm used.
Conclusion
RMA generally yields the most reproducible results, but we could not verify most genes identified as having changed expression by RMA with Real Time RT-PCR.
Supporting Evidence
- Calculating gene expression with six different algorithms resulted in different calculated gene expression levels.
- There was only 27 to 36% overlap between the different methods.
- RMA showed the highest correlation coefficients with Real Time RT-PCR data.
Takeaway
Different methods for measuring gene expression can give very different results, so it's important to choose the right one for your experiment.
Methodology
The study compared six algorithms (MAS5, dChip PMMM, dChip PM, RMA, GC-RMA, and PDNN) using Arabidopsis data to evaluate gene expression levels.
Limitations
The study could not verify most genes identified as differentially expressed by RMA with Real Time RT-PCR.
Participant Demographics
Arabidopsis thaliana accession Columbia-0.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website