AcconPred: Predicting Solvent Accessibility and Contact Number Simultaneously by a Multitask Learning Framework under the Conditional Neural Fields Model
Table 1
Precision, recall, and score for different evaluation dataset of 3-state solvent accessibility prediction.
Evaluation dataset
Precision
Recall
score
†Buried overall
0.76
0.78
0.77
‡Buried >0.9
0.96
0.31
0.47
Buried >0.8
0.92
0.45
0.60
Buried >0.7
0.88
0.57
0.69
Buried >0.6
0.84
0.66
0.74
Buried >0.5
0.79
0.74
0.76
Buried >0.4
0.75
0.82
0.78
Intermediate overall
0.56
0.50
0.53
Intermediate >0.9
1.00
0.0001
0.002
Intermediate >0.8
0.82
0.006
0.01
Intermediate >0.7
0.74
0.06
0.11
Intermediate >0.6
0.67
0.19
0.30
Intermediate >0.5
0.61
0.38
0.47
Intermediate >0.4
0.55
0.61
0.58
Exposed overall
0.71
0.76
0.73
Exposed >0.9
0.94
0.11
0.20
Exposed >0.8
0.88
0.31
0.46
Exposed >0.7
0.83
0.47
0.60
Exposed >0.6
0.78
0.61
0.68
Exposed >0.5
0.74
0.72
0.73
Exposed >0.4
0.69
0.81
0.75
Overall indicates the whole set of the predicted labels. ‡>0.9 indicates that the set of the predicted labels is chosen according to the predicted probability which is larger than 0.9.