Artificial Intelligence as a Catalyst for Antimicrobial Discovery: From Predictive Models to De Novo Design
Top Cited Papers
Open Access
- 6 February 2026
- journal article
- review article
- Published by MDPI AG in Microorganisms
- Vol. 14 (2) , 394
Abstract
Antimicrobial resistance represents one of the most critical global health challenges of the 21st century, urgently demanding innovative strategies for antimicrobial discovery. Traditional antibiotic development pipelines are slow, costly, and increasingly ineffective against multidrug-resistant pathogens. In this context, recent advances in artificial intelligence have emerged as transformative tools capable of accelerating antimicrobial discovery and expanding accessible chemical and biological space. This comprehensive review critically synthesizes recent progress in AI-driven approaches applied to the discovery and design of both small-molecule antibiotics and antimicrobial peptides. We examine how machine learning, deep learning, and generative models are being leveraged for virtual screening, activity prediction, mechanism-informed prioritization, and de novo antimicrobial design. Particular emphasis is placed on graph-based neural networks, attention-based and transformer architectures, and generative frameworks such as variational autoencoders and large language model-based generators. Across these approaches, AI has enabled the identification of structurally novel compounds, facilitated narrow-spectrum antimicrobial strategies, and improved interpretability in peptide prediction. However, significant challenges remain, including data scarcity and imbalance, limited experimental validation, and barriers to clinical translation. By integrating methodological advances with a critical analysis of the current limitations, this review highlights emerging trends and outlines future directions aimed at bridging the gap between in silico discovery and real-world therapeutic development.Keywords
This publication has 77 references indexed in Scilit:
- Discovery of antimicrobial peptides in the global microbiome with machine learningCell, 2024
- AMPActiPred: A three‐stage framework for predicting antibacterial peptides and activity levels with deep forestProtein Science, 2024
- Explainable AI in drug discovery: self-interpretable graph neural network for molecular property prediction using concept whiteningMachine Learning, 2023
- AMP‐BERT: Prediction of antimicrobial peptide function based on a BERT modelProtein Science, 2022
- Design of antimicrobial and cytolytic peptides by computational analysis of bacterial, algal, and invertebrate proteomesAmino Acids, 2020
- A Deep Learning Approach to Antibiotic DiscoveryCell, 2020
- Antimicrobial peptides (AMPs): Ancient compounds that represent novel weapons in the fight against bacteriaBiochemical Pharmacology, 2017
- Beta-Lactam Antibiotics Induce a Lethal Malfunctioning of the Bacterial Cell Wall Synthesis MachineryCell, 2014
- Molecular understanding of aminoglycoside action and resistanceApplied Microbiology and Biotechnology, 2006
- Experimentelle Grundlage der Chemotherapie der Spirillosen1910