Fig 1.
Overview of the CMCL-DDI framework.
It comprises three main components: (a) graph view module, which encodes pharmacophore-level structural features using a Transformer-based encoder and readout function. Each pharmacophore box corresponds to the subgraphs decomposed from a single molecule; (b) sequence view module, which extracts semantic representations from SMILES strings using MOLBERT; and (c) cross-view contrastive learning module, which aligns the representations from both views during training. (d) DDI prediction module, which fuses the learned representations via a cross-attention mechanism and predicts interaction probability using a multi-layer perceptron.
Fig 2.
The architecture of transformer.
Table 1.
Hyperparameter configurations of model experiments.
Table 2.
Confusion matrix for prediction results.
Table 3.
The performance of CMCL-DDI and baselines on two datasets in the warm-start setting (%).
Fig 3.
Performance comparison of different models under warm-start setting.
The left figure displays results on the DrugBank dataset, and the right figure shows results on the Twosides dataset.
Table 4.
The performance of CMCL-DDI and baselines on two datasets in the cold-start setting (%).
Fig 4.
Performance comparison of different models under cold-start setting.
The left figure displays results on the DrugBank dataset, and the right figure shows results on the Twosides dataset.
Table 5.
Ablation study performance of CMCL-DDI.
Fig 5.
Sensitivity analysis on the dimension of feature embeddings.
Fig 6.
Sensitivity analysis of the number of attention heads.
Fig 7.
Case studies demonstrating pharmacophores identified by CMCL-DDI in clinically confirmed DDIs.