Computer-Aided Design of Antimicrobial Peptides: Are We Generating Effective Drug Candidates?

Abstract

Computer-Aided Design of Antimicrobial Peptides http://crdd.osdd.net/raghava/antibp2/ ClassAMP Uses RF and SVM to predict the propensity of a protein sequence to have antibacterial, antifungal, or antiviral activity http://www.bicnirrh.res.in/classamp/ AMPA Web tool for assessing the antimicrobial domains of proteins, with a focus on the design on new antimicrobial drugs http://tcoffee.crg.cat/apps/ampa/do DBAASP Provides users with information on detailed structure (chemical, 3D) and activity for those peptides, for which antimicrobial activity against particular target species have been evaluated experimentally https://dbaasp.org/home Joker An algorithm to design antimicrobial peptides using their language https://github.com/williamfp7/Joker

Bioinformatic Analysis of 1000 Amphibian Antimicrobial Peptides Uncovers Multiple Length-Dependent Correlations for Peptide Design and Prediction

“…Using a computing approach, Juretić and colleagues designed a selective peptide [ 156 ]. Quantitative structure–activity relationship (QSAR) studies are also applied to the design of antimicrobial and antibiofilm peptides [ 157 , 158 ]. We developed database-guided methods [ 159 ].…”

Antimicrobial peptides as therapeutic agents: opportunities and challenges

“…Importantly, several computational tools have recently been developed to assist in designing and optimizing AMP variants with improved efficacy. Among them, machine-learning methods are frequently used, with the focus on a quantitative structure-activity relationship (QSAR) model, which applies physicochemical descriptors to predict the biological efficacy of peptides from their amino acid sequences [105]. In addition, the de novo computational methods, generating AMP sequences without a seed sequence but using amino acid frequency/position preferences to guarantee certain characteristics, such as load, amphipathicity, and structure, have allowed the production of a large number of peptide variants with a great diversity in amino acid composition as well as 3D structure [105].…”

“…Among them, machine-learning methods are frequently used, with the focus on a quantitative structure-activity relationship (QSAR) model, which applies physicochemical descriptors to predict the biological efficacy of peptides from their amino acid sequences [105]. In addition, the de novo computational methods, generating AMP sequences without a seed sequence but using amino acid frequency/position preferences to guarantee certain characteristics, such as load, amphipathicity, and structure, have allowed the production of a large number of peptide variants with a great diversity in amino acid composition as well as 3D structure [105]. Evolutionary algorithms, particularly genetic algorithms, in which the process of evolution is performed in silico through successive generations of mutations and deletions, are also broadly applied to search for AMPs with improved activity [105].…”

“…In addition, the de novo computational methods, generating AMP sequences without a seed sequence but using amino acid frequency/position preferences to guarantee certain characteristics, such as load, amphipathicity, and structure, have allowed the production of a large number of peptide variants with a great diversity in amino acid composition as well as 3D structure [105]. Evolutionary algorithms, particularly genetic algorithms, in which the process of evolution is performed in silico through successive generations of mutations and deletions, are also broadly applied to search for AMPs with improved activity [105]. Another main strategy to improve the half-life and performance of AMPs is to use advanced delivery vehicles.…”

Strategies in Translating the Therapeutic Potentials of Host Defense Peptides

“…Long been deployed in the automobile and technology industries, AI has only started gaining traction in the field of science and medicine, owing to the advancement in computer power, availability of big data, publicly available neural networks, and improvement in AI algorithms using machine learning and deep learning (177)(178)(179)(180). In view of the infinite chemical space and complex SAR of natural and synthetic HDPs, AI serves as an attractive solution to identify and predict novel peptide sequences with potentially good antimicrobial efficacy (181)(182)(183)(184).…”