Resistance Peptide Identification From Fish And Penaeus Based On Amino Acid Composition And Physicochemical Properties

Resistant genes and animal immune activity factors play an significant role in the immune system of animals and plants.They are often used as selective genetic markers to develop excellent stress-resistance trait lines and can be well used in the cultivation of Larimichthys crocea.Antimicrobial peptides are important for the sustainable development of fish and penaeus aquaculture due to their non-resistance and their unique mechanism of action and their multi-effect antibacterial activity,and are considered to be a substitute for antibiotics.Therefore,how to efficiently identify resistant genes and multi-responsive antimicrobial peptides has become a top priority of research.With the rapid development of bioinformatics technology,the application of computational methods makes the mining work more effective in the face of large-scale resistant polypeptide sequences,and different feature algorithms and statistical analysis are used to mine the correlation and differences between sequences.In this regard,resistant peptides of fish and penaeus were taken as the research object,and the primary research will be implemented in the following aspects:First of all,based on immune similarity between animals and plants,this paper proposes for the first time that prediction of potential resistance genes in Larimichthys crocea by constructing a dataset containing plant resistance genes and animal immune related factors data sets,which can help us to learn about the immune mechanism of Larimichthys crocea deeply.Secondly,Based on the 188-dimensional features of amino acid composition and physicochemical property extraction sequences,a large yellow croaker resistance gene prediction model was constructed in combination with the random forest algorithm,and an online prediction system for resistant genes was implemented.In addition,in this paper,a two-stage classification model is proposed for the identification of antimicrobial peptides and their antimicrobial functions using the amino acid composition and physico-chemical features extraction method.The first classifier identifies the tested peptides,and the second-stage classifiers the function of pleiotropic antimicrobial peptides.For the purpose of verifing the validity of this classification model,this article compares with the existing literature methods.At the same time,the peptide sequence of penaeus was predicted by using this two-stage classification model.This paper analyzes and evaluates the combined classifiers of multiple feature extraction algorithms and classification algorithms.The results show that the proposed 188D-RF classification algorithm can effectively classify resistance genes and antimicrobial peptides.The PS-RF classification algorithm can accurately classify multi-label antimicrobial peptides and achieve good expectations.At the same time,the classification work done in our experiments can be extended to other biological fields.