Fig. 4From: MCL-DTI: using drug multimodal information and bi-directional cross-attention learning method for predicting drug–target interactionArchitectural elements of a cross attention block between two time-seriers form drug \(\alpha \) and target \(\beta \)Back to article page