IAMPE: NMR-Assisted Computational Prediction of Antimicrobial Peptides

Abstract

Antimicrobial peptides (AMPs) are at the focus of attention due to their therapeutic importance and developing computational tools for the identification of efficient antibiotics from the primary structure. Here, we utilized the 13CNMR spectral of amino acids and clustered them into various groups. These clusters were used to build feature vectors for the AMP sequences based on the composition, transition, and distribution of cluster members. These features, along with the physicochemical properties of AMPs were exploited to learn computational models to predict active AMPs solely from their sequences. Naïve Bayes (NB), k-nearest neighbors (KNN), support-vector machine (SVM), random forest (RF), and eXtreme Gradient Boosting (XGBoost) were employed to build the classification system using the collected AMP datasets from the CAMP, LAMP, ADAM, and AntiBP databases. Our results were validated and compared with the CAMP and ADAM prediction systems and indicated that the synergistic combination of the 13CNMR features with the physicochemical descriptors enables the proposed ensemble mechanism to improve the prediction performance of active AMP sequences. Our web-based AMP prediction platform, IAMPE, is available at .

Computational Methods and Tools in Antimicrobial Peptide Research

“…113 A more recent comparison published in the last year by the developers of the AmpGram prediction tool 114 naturally suggests that their own method is more accurate, but it should be noted that the different analyzed predictors, which include CAMPR3, AMPScanner, ADAM, iAMPpred, and iAMP-2L, have largely yielded high performances across different data sets. The most recent review of AMP prediction tools available in the literature 115 reaches a similar conclusion, with several tools having comparable accuracy; the ones that performed the best, across multiple data sets, were amPEPpy 116 and AMPfun, 117 although, on a set of ten more recent AMPs, IAMPE 118 had the highest correct prediction rate.…”

An Efficient Evaluation System Accelerates α-Helical Antimicrobial Peptide Discovery and Its Application to Global Human Genome Mining

“… Liu et al (2018) established an activity prediction method based on the predicted 3D descriptors of AMP, in which, the antibacterial effect of the novel AMPs designed by this method was from 32 to 512 μg/ml. Kavousi et al (2020) utilized the 13 C-NMR spectral of amino acid combined with the physicochemical properties of AMPs, such as amino acid acidity and basicity, size, charged percentages, and so on, to establish a computational approach to predict AMPs, and the result showed a 95% accuracy but no detailed antimicrobial activity released. The artificial intelligence method reported by Torres et al (2021) to predict AMP from human proteome showed a 63.6% hit rate, and 55 synthesized representative peptides showed MIC values from 0.39 to 128 μg/mL against pathogens.…”

A computational learning paradigm to targeted discovery of biocatalysts from metagenomic data: A case study of lipase identification

“…In some instances, sequence similarity might not directly correlate with the resemblance of various enzymatic properties. In such cases, machine learning (ML) approaches could facilitate the in silico characterization of target enzymes before engaging wet-lab experiments (Atanaki et al, 2020;Kavousi et al, 2020;Shahraki et al, 2019bShahraki et al, , 2020. Recently, data science revolutions have a great effect on conventional enzyme research and could speed up enzyme discovery procedure.…”

“…Recently, data science revolutions have a great effect on conventional enzyme research and could speed up enzyme discovery procedure. ML may play a central role in the paradigm change away from using traditional methods in different applications such as peptide function prediction (Atanaki et al, 2020;Kavousi et al, 2020;Nasiri et al, 2021), enzyme characteristics prediction (Ariaeenejad et al, 2018(Ariaeenejad et al, , 2021Shahraki et al, 2019aShahraki et al, , 2020, and so on. In a review done by Toyao et al (Toyao et al, 2020), recent developments have been reported in the use of ML to build homogeneous and heterogeneous catalysts.…”