| Dataset | Organism | Platform | Features | Expected DEG | Negative spots | Replaced |
---|
A | Baird et al. | Homo sapiens | cDNA | 12,600 | 2,045 (0.16) | 7,444 (0.59) | With zero |
B | Urban et al. | S. Cerevisiae | Oligo | 10,789 | 3,719 (0.34) | None | - |
C | Smith et al. | S. Cerevisiae | Oligo | 25,240 | 27 (0.004) | 524 (0.08) | No |
D | De Pittà et al. | Homo sapiens | cDNA | 9,984 | 1,353 (0.13) | None | - |
- For each dataset the table reports the organism, the platform, the number of features, the number of expected differentially expressed genes (DEG) (in brackets the expected proportion), the mean number of negative spots per array (in brackets the mean proportion), and if the negative spots are replaced or not. Note that in dataset C each array contains four replicates of each gene, giving a total of 6,130 unique features. Therefore, the proportion of genes expected to be differentially expressed has been calculated on the total of unique features. Expected number of DEG is calculated as the average number of differentially expressed genes obtained with SAM test after all the normalisations used in the study.