Supporting information for Qian et al. (2002) Proc. Natl. Acad. Sci. USA, 10.1073/pnas.222491799
Table 4. Summary of the comparison of results of four replicate experiments
Comparison | No change call, % |
1 vs. 2 | 99.7 |
1 vs. 3 | 99.5 |
1 vs. 4 | 98.7 |
2 vs. 3 | 99.7 |
2 vs. 4 | 98.6 |
3 vs. 4 | 99.1 |
Average | 99.1 |
To detect experimental variation derived from cDNA preparation, cRNA labeling, hybridizations, and washing conditions, we tested the reproducibility of our microarray assay by repeating the procedure four times with total RNA isolated from the KG-1 myeloid leukemia cell line. The Affymetrix software was used to perform comparison analyses. The left column shows all of the possible comparisons among the four expression profiles, and the right column shows the percentage of "no change call," which is defined by the algorithm within the Affymetrix software. "No change call" is a measure of the genes/ESTs that do not change in their expression levels between two samples and is calculated from intensity readings of genes/ESTs, background intensity, and other factors. Overall, we saw excellent reproducibility, because our average "no change call" was above 99%.