Table 3.
Module classifier performance Download Figure 4-1, TIF file.
| Comparison | Modules (number) | Cv error (%) | Cv error Std | Mc error (%) | Mc EP | ||
|---|---|---|---|---|---|---|---|
| Wild type | |||||||
| •Day/Night | 10 | 1.61 | 1.29 | 50.0 | 2.25 | ||
| •Day 5/Day 6 | 8 | 21.0 | 6.53 | 50.0 | 3.18 | ||
| •Night 5/Night 6 | 1 | 35.5 | 9.71 | 50.0 | 3.18 | ||
| hcrtr | |||||||
| Day and night | |||||||
| •WT/Het | 1 | 27.7 | 0.77 | 27.7 | 1.88 | ||
| •WT/Hom | 10 | 45.8 | 10.9 | 50.0 | 2.83 | ||
| •Het/Hom | 8 | 27.5 | 1.12 | 27.7 | 1.88 | ||
| Day | |||||||
| •WT/Het | 1 | 27.7 | 1.46 | 27.7 | 2.66 | ||
| •WT/Hom | 1 | 40.4 | 12.5 | 50.0 | 4.00 | ||
| •Het/Hom | 3 | 27.3 | 2.35 | 27.7 | 2.66 | ||
| Night | |||||||
| •WT/Het | 1 | 27.7 | 1.46 | 27.7 | 2.66 | ||
| •WT/Hom | 1 | 47.4 | 10.9 | 50.0 | 4.00 | ||
| •Het/Hom | 10 | 27.0 | 1.72 | 27.7 | 2.66 | ||
| Melatonin (day) | |||||||
| •Control | 3 | 8.33 | 8.69 | 25.0 | 4.42 | ||
| •0.01μM | 10 | 2.78 | 6.02 | 16.7 | 4.39 | ||
| •0.1μM | 2 | 16.7 | 4.52 | 16.7 | 4.39 | ||
| •1μM | 1 | 18.1 | 7.74 | 16.7 | 4.39 | ||
| •3μM | 1 | 16.8 | 8.67 | 16.7 | 4.39 | ||
| •10μM | 1 | 16.8 | 4.52 | 16.7 | 4.39 | ||
| •30μM | 1 | 16.8 | 4.52 | 16.7 | 4.39 | ||
| PTZ (day) | |||||||
| •Control | 1 | 1.92 | 5.27 | 46.2 | 6.91 | ||
| •2.5 mM | 1 | 17.9 | 17.6 | 35.7 | 9.06 | ||
| •5 mM | 1 | 28.6 | 22.3 | 32.1 | 8.83 | ||
| •7.5 mM | 10 | 20.0 | 26.1 | 32.1 | 8.83 |
A table showing the performance of each module classifier. Each classifier sought to separate the data shown in the comparison column, e.g., wild type, day/night. For each comparison, all 10 modules were sequentially chosen by the mRMR algorithm, then a smaller subset was retained (see module column) based on classification error curves. Cv, 10-fold cross validated; Std, SD across the 10 folds; Mc, majority class classifier; EP, SE of proportion.