TABLE I.
Population Statistics for Our Datasets and Tasks.
| Dataset Statistics: | ||||
|---|---|---|---|---|
| Dataset | Number of Samples | # Cell Lines | Most Frequent Cell Line | Least Frequent Cell Line |
| Full LINCS | 156,461 | 36 | MCF7 (26,546) | NCIH716 (8) |
| Prostate Only LINCS | 25,565 | 2 | PC3 (13,625) | VCAP (11,940) |
| MGH NeuroBank (Level 4) | 5602 | 5 | N/A (1133) | N/A (1109) |
| MGH NeuroBank (Level 5) | 1894 | 5 | N/A (380) | N/A (377) |
| Task Statistics: | ||||
|---|---|---|---|---|
| Dataset | Task | # Classes | Most Frequent Class | Least Frequent Class |
| LINCS (Full) | Primary Site | 12 | Prostate (43,686) | Ovary (415) |
| Subtype | 14 | Adenocarcinoma (53,245) | Embryonal Kidney (1384) | |
| MOA | 49 | DMSO (25,638) | IKK Inhibitor (828) | |
| LINCS (Prostate Only) | MOA | 9 | DMSO (8833) | Serotonin Receptor Antagonist (1029) |
| MGH NeuroBank (Level 4) | Perturbagen | 60 | DMSO (383) | Ruboxistaurin (78) |
| MGH NeuroBank (Level 5) | Perturbagen | 60 | DMSO (130) | Ruboxistaurin (27) |