Table 1:
Assessment of accuracy and precision based on the Affymetrix HGU133A spike-in data
| Precision | Accuracy | |
|---|---|---|
| Within-replicate variance, mean (SD) | Non-linearity, mean (SD) | |
| MBEI—no MM | 0.030 (0.020) | 9.102 (2.416) |
| PLIER—no MM | 0.068 (0.057) | 9.895 (2.787) |
| RMA—no BG | 0.093 (0.062) | 9.713 (2.372) |
| fRMA—default | 0.097 (0.071) | 6.954 (2.558) |
| RMA—default | 0.148 (0.072) | 7.401 (2.909) |
| MBEI—default | 0.181 (0.121) | 5.664 (3.058) |
| GCRMA—default | 0.251 (0.170) | 5.647 (3.150) |
| MAS5.0—default | 0.991 (0.993) | 4.423 (2.924) |
| PLIER—default | 1.849 (3.155) | 4.809 (3.307) |
For each probe set, the within-replicate variance averaged across nominal concentrations is used as a measure of precision, and the nonlinearity of the signal detect curve is used as a measure of accuracy. The mean and SD across probe sets is reported for each preprocessing algorithm.
Note: MM, mismatch; BG, background correction.