Table 1.
Precision on the Selection of the Differentially Expressed Genes Using Raw Data or Preprocessed Data with RMA and MAS5
| Group comparison | raw | RMA | MAS5 |
|---|---|---|---|
| 1 vs. 2 | 7.14 | 9.52 | 4.76 |
| 1 vs. 3 | 26.19 | 16.67 | 16.67 |
| 1 vs. 4 | 38.10 | 11.90 | 14.29 |
| 1 vs. 5 | 28.57 | 28.57 | 16.67 |
| 1 vs. 6 | 26.19 | 28.57 | 28.57 |
| 1 vs. 7 | 40.48 | 26.19 | 23.81 |
| 1 vs. 8 | 35.71 | 21.43 | 30.95 |
| 1 vs. 9 | 40.48 | 23.81 | 23.81 |
| 1 vs. 10 | 35.71 | 19.05 | 21.43 |
| 1 vs. 11 | 38.10 | 14.29 | 21.43 |
| 1 vs. 12 | 23.81 | 16.67 | 9.52 |
| 1 vs. 13 | 23.81 | 23.81 | 14.29 |
| 1 vs. 14 | 7.14 | 4.76 | 9.52 |
| Mean precision | 28.57 | 18.86 | 18.13 |
The data are the Affymetrix Latin Square Data for Expression Algorithm Assessment. The selection is performed between the first group and the rest to include all the differences among the spike-in concentrations. The best result for each case (raw data, RMA preprocessed data, and MAS5 preprocessed data) is shown in boldface.
RMA, Robust Multiarray Average.