A Deep Convolutional Neural Network-Based Method for Predicting DNA-Binding Residues

Abstract

A method for predicting DNA binding residues based on a deep convolutional neural network. First, according to the input protein sequence information whose number of residues is L to be predicted for ligand binding residues, use the psi-blast program and the PSSpred program to obtain matrix PSSM and PSS; then, combine the two matrices into a feature matrix F; secondly, we process the protein sequence into residue samples; thirdly, build a deep convolutional neural network, using the protein sequence of known binding residues to construct Data set, and the data set is divided into M groups of data subsets, using these ten sets of data subsets to train M network models; finally, the protein sequence to be predicted is processed into residue samples, and input to the trained In the M network models of , the prediction results of these M models are combined to predict whether the residues in the protein sequence are binding residues. The invention has low calculation cost and high prediction accuracy.

Description

technical field [0001] The invention relates to the fields of bioinformatics, pattern recognition and computer applications, in particular to a method for predicting DNA binding residues based on a deep convolutional neural network. Background technique [0002] Protein-ligand interaction is ubiquitous and indispensable in life processes, and this interaction plays a very important role in the recognition and signal transmission of biomolecules. Among them, DNA molecules belong to the category of ligand molecules. Accurately identifying the binding residues of DNA molecules in protein sequences is helpful for understanding protein functions, analyzing the interaction mechanism between proteins and DNA molecules, and designing drug target proteins. important biological significance. [0003] Research literature found that many methods for predicting DNA-binding residues in protein sequences have been proposed, such as: DISPLAR (Tjong H, Zhou H.an accurate method for predicti...

AI Technical Summary

Problems solved by technology

Although existing methods can be used to predict DNA-binding residues in protein sequences, a large amount of experimental data and machine learning algorithms are generally used, so the cost is high. The accuracy is not guaranteed to be optimal and needs to be further improved

[0004] In summary, the existing methods for predicting DNA binding residues are still far from the requirements of practical applications in terms of computational cost and prediction accuracy, and urgently need to be improved.

Images