Fig. 4
From: CNN-Siam: multimodal siamese CNN-based deep learning approach for drug‒drug interaction prediction
The framework of CNN-Siam. a Workflow: Inputting Drug A and Drug B into two weight-sharing CNNs to obtain the feature outputs, summing them, and finally inputting them into a multilayer perceptron for classification. b Model architecture of the CNN. It consists of 5 convolutional layers and 1 normalization layer, where the output of the 2nd convolutional layer is added to the input of the 5th convolutional layer to achieve a residual connection. c Architecture of the MLP. Two hidden layers and one classification layer