EBAMP: An efficient de novo broad-spectrum antimicrobial peptide discovery framework
Yunxiang Zhao, Qian Li, Mingyue Sun, Yixin Su, Xinxin Su, Ling Jiang, Xinran Zhang, Yida David Hu, Boqian Wang, Haoran Yu, Qiang Zhang, Zili Chai, Yao Ding, Yuhao Ren, Wenhua Huang, Yuan Jin, Yutao Dou, Dongsheng Li, Zhen Huang, Pengpeng Liu, Hongguang Ren, Yongqiang Jiang
Abstract
De novo design of antimicrobial peptides (AMPs) is challenging due to the vast combinatorial space and unknown mechanisms. We propose EBAMP, a generative-discriminative framework for de novo broad-spectrum AMP design targeting bacteria and fungi. EBAMP combines a Transformer-based generative model with advanced feature-based screening to explore peptide space and select multiobjective candidates. Experimental testing of 256 designed sequences shows that 96 (37.5%) display bactericidal ability. The top 10 sequences exhibit low cytotoxicity, low hemolysis, and strong antibacterial effect (2 μg/mL) against multidrug-resistant bacteria and fungi. In vivo mouse full-thickness wound infection model demonstrates inhibitory effects against Acinetobacter baumannii (bacterium) and Candida auris (fungus), with therapeutic efficiency comparable to antibiotics but lower resistance propensity. Alanine substitution analysis and molecular dynamics reveal functionally critical positions. EBAMP showcases large generative models for broad-spectrum AMP discovery and addresses antibiotic resistance.