Table 1 Classification performance of evaluated methods for CPI prediction
From: GCRNN: graph convolutional recurrent neural network for compound–protein interaction prediction
Method | Precision | Recall | Accuracy | F1-measure |
|---|---|---|---|---|
GCNN | 0.92 | 0.94 | 0.96 | 0.92 |
GCRNN w/LSTM | 0.95 | 0.92 | 0.97 | 0.93 |
GCRNN w/GRU | 0.97 | 0.94 | 0.98 | 0.95 |