Prediction of Linear Cationic Antimicrobial Peptides Active against Gram-Negative and Gram-Positive Bacteria Based on Machine Learning Models

Abstract

Antimicrobial peptides (AMPs) are considered as promising alternatives to conventional antibiotics in order to overcome the growing problems of antibiotic resistance. Computational prediction approaches receive an increasing interest to identify and design the best candidate AMPs prior to the in vitro tests. In this study, we focused on the linear cationic peptides with non-hemolytic activity, which are downloaded from the Database of Antimicrobial Activity and Structure of Peptides (DBAASP). Referring to the MIC (Minimum inhibition concentration) values, we have assigned a positive label to a peptide if it shows antimicrobial activity; otherwise, the peptide is labeled as negative. Here, we focused on the peptides showing antimicrobial activity against Gram-negative and against Gram-positive bacteria separately, and we created two datasets accordingly. Ten different physico-chemical properties of the peptides are calculated and used as features in our study. Following data exploration and data preprocessing steps, a variety of classification algorithms are used with 100-fold Monte Carlo Cross-Validation to build models and to predict the antimicrobial activity of the peptides. Among the generated models, Random Forest has resulted in the best performance metrics for both Gram-negative dataset (Accuracy: 0.98, Recall: 0.99, Specificity: 0.97, Precision: 0.97, AUC: 0.99, F1: 0.98) and Gram-positive dataset (Accuracy: 0.95, Recall: 0.95, Specificity: 0.95, Precision: 0.90, AUC: 0.97, F1: 0.92) after outlier elimination is applied. This prediction approach might be useful to evaluate the antibacterial potential of a candidate peptide sequence before moving to the experimental studies.

Bacteria-Specific Feature Selection for Enhanced Antimicrobial Peptide Activity Predictions Using Machine-Learning Methods

“…A major drawback in previous machine-learning predictions of antimicrobial peptide activity is that those models are not bacteria specific. A recent study tried to circumvent this problem by predicting the antimicrobial activity against Gram-negative and Gram-positive bacteria and utilized data sets for peptide activity against three Gram-negative bacteria, E. coli and A. baumannii , and P. aeruginosa . Although these species share some common membrane architecture, there are key differences between them that might affect antimicrobial activity against them .…”

Predicting Antimicrobial Activity for Untested Peptide-Based Drugs Using Collaborative Filtering and Link Prediction

“…S1. The largest number of AMP neighbors were for the antibiotics rifampicin (17), erythromycin (12), and kanamycin (8), and these antibiotics also had the largest number of common neighbors: 4 for rifampicin and kanamycin, 4 for rifampicin and erythromycin, and 5 for erythromycin and kanamycin. Thus, it is not surprising that by preferential attachment (see Fig.…”

“…The data are even sparser for AMP-AMP combinations, which may contribute to the lack of studies on elucidating molecular mechanisms of antibacterial action for existing AMP-AMP combinations and predicting activity of new ones [16]. Previous studies focused on the efficacy of single AMP types alone and were limited in combining very different bacteria for their analyses (e.g., broad categories of gram-positive vs gram-negative or active against one type vs inactive against all tested types [17,18]). Although certain AMP types and AMP combinations have shown broad-spectrum antimicrobial activity, there are key differences between bacteria, for example in membrane composition [19] or morphology [20], that can affect the efficacy of a single AMP type or AMPs in combination and render some bacteria but not others resistant to November 17, 2022 2/21 the antimicrobial agents.…”

Prediction of the Antibacterial Activity of the Green Synthesized Silver Nanoparticles against Gram Negative and Positive Bacteria by using Machine Learning Algorithms

“…For instance, Tiihonen et al [34] studied the antimicrobial activity of conjugated oligoelectrolyte molecules against Escherichia coli. Söylemez et al [35] also utilized a ML for prediction of antimicrobial activity of peptides against gram negative and positive bacteria. Her and Wu [36] explored using a pan-genome-based ML method for prediction of the antimicrobial resistance activities of E. coli.…”