An optimal multi-marker ensemble prediction method for multifunctional antimicrobial peptides

Abstract

The invention belongs to the field of bioinformatics and relates to an optimal multi-tag integrated prediction method for multifunctional anti-microbial peptide. The method comprises the steps of extracting various pseudo amino acid composition eigenvectors, with different dimensions, of the anti-microbial peptide by adopting a pseudo amino acid composition quantification method based on amino acid sequence information of the anti-microbial peptide; training out a plurality of different multi-tag classifiers on all the generated pseudo amino acid composition eigenvectors with the different dimensions by applying a multi-tag classification algorithm; screening out an optimal multi-tag classifier subset for each function type of the anti-microbial peptide from the obtained multi-tag classifiers; and inputting a to-be-predicted peptide sequence in the classifier subset, and executing a weighted majority voting fusion policy to obtain attributions of N function types of the peptide sequence. According to the method, an optimal characteristic parameter combination does not need to be determined in advance, so that the situation that the optimal characteristic parameter combination is selected through a large amount of traversal experiments during modeling can be avoided, the practicality of the prediction method is improved, and the modeling efficiency is enhanced.

Description

technical field [0001] The invention belongs to the field of bioinformatics, relates to the field of function prediction of antimicrobial peptides, and in particular relates to an optimal multi-label integrated prediction method of multifunctional antimicrobial peptides. Background technique [0002] Antimicrobial peptides, also known as host defense peptides, are an important class of biomacromolecules in the innate immune system of organisms. They are present in almost all living organisms and protect the organism itself from infection by pathogenic bacteria. Antimicrobial peptides have natural immune properties and are excellent alternatives to traditional antibiotic drugs to address the problem of antibiotic resistance. [0003] With the generation of a large number of protein sequences in the post-genomic era, the gap between known sequences of antimicrobial peptides and unknown protein sequences is increasing. Experimentally identifying which protein sequences are an...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide an optimal multi-marker integrated prediction method for multifunctional antimicrobial peptides, aiming to solve the problems of poor prediction effect and low modeling efficiency of existing prediction methods for multifunctional antimicrobial peptides

Images