1- Collection of all the verified sulfation sites of human proteins from UniProt, and then pre-process them by removing redundancy.
3 - Attempting to represent protein sequences as continuous vectors (embeddings) by using the language model ELMo taken from natural language processing.
4 - Evaluate the predication performance and compare it to the other known methods.
5 - Deploying the model in the web using Streamlit and AWS.