Evaluation

Plotting evaluation metrics, such as ROCPR, are important for understanding predictive ability. In the current version, we can use PyPropel to draw ROCPR curves.

We use interaction predicted probabilities by the tma300 tool.

tma300_roc_fpr_custom.json contains false positive rate (FPR) and tma300_roc_tpr_custom.json contains true positive rate (TPR).

Python 

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
import pypropel as pp

X_fpns = {
    'tma300': to('data/eval/tma300/tma300_roc_fpr_custom.json'),
}
Y_fpns = {
    'tma300': to('data/eval/tma300/tma300_roc_tpr_custom.json'),
}

fig, ax = plt.subplots(
    nrows=2,
    ncols=2,
    # figsize=(6, 5),
    figsize=(12, 10),
    sharey='all',
    sharex=False,
)
pp.plot.rocpr(
    X_fpns,
    Y_fpns,
    x_label='fpr',
    y_label='tpr',
    title='',
    ax=ax[0, 0],
)
pp.plot.rocpr(
    X_fpns,
    Y_fpns,
    x_label='fpr',
    y_label='tpr',
    title='',
    ax=ax[0, 1],
)
pp.plot.rocpr(
    X_fpns,
    Y_fpns,
    x_label='fpr',
    y_label='tpr',
    title='',
    ax=ax[1, 0],
)
plt.show()

Image title
Fig 1. ROCPR curves