Table 2.
State of the art sequence-based PPI predictors.
| Predictor | Type | Machine learning model | Features | Availability | Ref. |
|---|---|---|---|---|---|
| SPRINT | Similarity-based scoring | Not applicable | Not applicable | C++ source | 15 |
| PIPR | Machine learning | RCNN | Latent representation of protein sequences generated by a combination of convolution layers and re- sidual gated recurrent units | Python source | 18 |
| D-SCRIPT | Machine learning | Bi-LSTM + CNN | Combination of a pre-trained embedding model with a predicted contact map | Python package | 24 |