Protein phosphorylation site prediction method based on inner product self-attention neural network

Abstract

The invention relates to a protein phosphorylation site prediction method based on an inner product self-attention neural network. The method comprises the steps that: firstly, a protein sequence which has an quantityL of amino acid residues and is to be subjected to phosphorylation site prediction is input, and the protein sequence is converted into an L * 20 feature matrix Mfea through one-hot coding expression forms of 20 common amino acids; then, for each residue of the protein, a characteristic matrix M with the size of w * 20 is acquired by using a sliding window according to the Mfea; protein sequences of known phosphorylation site tags are obtained from a database to construct a training set; a neural network framework is built, the network is trained by using the training set, and a prediction model is obtained; and finally, the feature matrix of the protein sequence residues to be predicted is input into the trained model, and whether target residues in the protein sequence are phosphorylation sites or not is predicted according to a probability value output by the model. The method is low in calculation cost and high in prediction accuracy.

Description

technical field [0001] The present invention relates to the field of bioinformatics, deep learning and computer application, in particular to a protein phosphorylation site prediction method based on inner product self-attention neural network. Background technique [0002] Protein phosphorylation is a post-translational modification process widely present in eukaryotes, which plays an important role in various biological processes such as energy metabolism, signal transduction pathways, neural activity, cell cycle and apoptosis. Accurate identification of protein phosphorylation sites can not only help us understand complex protein biological systems, but also guide the design of basic biomedical drugs. [0003] Currently, methods for predicting protein phosphorylation sites through deep learning include: NetPhos3.0 (Blom, N. et al. Prediction of post-translational glycosylation and phosphorylation of proteins from the amino acid sequence [J]. Proteomics, 2004: 4 , 1633–16...

AI Technical Summary

Problems solved by technology

[0004] In summary, the existing protein phosphorylation site prediction methods are still far from the requirements of practical application in terms of calculation cost and prediction accuracy, and urgently need to be improved.

Images