Prediction of Phosphorylation Sites in Amino Acid Sequences Using Convolutional Neural Networks

Abstract

The focus of this research project is to predict the location of phosphorylation on human proteins by their amino acid sequences. Signals are sent into cells through receptors during a process called signaling cascade, which may alter the state of cells. One essential element of signaling cascade is protein kinases, which are enzymes that give the target protein a phosphate group. They can activate, deactivate proteins or act as signals to be passed onto other proteins. Through the Convolutional Neural Network (CNN), our model aims to provide a plausible prediction of the phosphorylation pattern of a given human protein sequence, on which future experiments can be based.