Fig. 5From: DeepNetBim: deep learning model for predicting HLA-epitope interactions based on network analysis by harnessing binding and immunogenicity informationThe architecture of the deep neural network in DeepNetBim. For an HLA-peptide pair, the encoded peptide (encoded to 9 × 21 matrix, coloured in blue) and network metrics (encoded to 1 × 8 vector, coloured in green) were fed into both convolutional and attention modules. The outputs of the two modules were then merged together by concatenating them to a single tensorBack to article page