Siamese networks for protein classification

Siamese networks are ANN architectures working in tandem on two different input vectors and using the same weights to compute comparable output vectors. This work explores how siamese networks can be used for deep representation learning classifying protein-protein interactions.