Table 1.
Descriptive statistics for the testing and predictive sets.
| Descriptives | Testing Set | Predictive Set | |
|---|---|---|---|
| PKI Group | AOD Group | ||
| No. of compounds | 9137 | 18 | 80 |
| No. of cell lines | 60 | 60 | 60 |
| Total datapoints | 548,220 | 1080 | 4800 |
| Missing values | 33,861 | 49 | 122 |
| Average no. of datapoints/compound | 56.29 | 57.28 | 58.47 |
| Average pGI50 value | 5.10 | 5.74 | 5.46 |
| Total no. of outliers | 36,570 | 69 | 253 |
| No. of outliers/compound | 1–22 (Avg* = 4.00) |
0–8 (Avg = 3.83) |
0–14 (Avg = 3.16) |
| Range* | 0.002–6.65 (Avg = 1.58) |
0.54–3.98 (Avg = 2.27) |
0.12–4.4 (Avg = 1.98) |
| Standard deviation | 0.0003–1.6617 (Avg = 0.3022) |
0.1168–1.3622 (Avg = 0.5116) |
0.028–0.9999 (Avg = 0.4242) |
No. = number; Avg* = average value; Range* (for a compound) = difference between the maximum value and minimum value for all of the compound’s datapoints; PKI = protein kinase inhibitors; AOD = approved oncologic drugs; pGI50 = negative log values of 50% growth inhibition concentration.