Fig. 1
From: DeepSuccinylSite: a deep learning based approach for protein succinylation site prediction
a Window size of 33 in FASTA format is the input. It is converted into integers which is then encoded either using one-hot encoding or embedding layer. This will be the input for CNN layers. b The output from either of the encoding is then fed as input into the deep learning architecture. Finally, after the flattening and fully connected layers we get the final output which contains two nodes with outputs [0 1] for positive and [1 0] for negative sites