计算生物学
生化工程
生物
抗菌肽
蛋白质组
合成生物学
生物技术
计算机科学
翻译(生物学)
抗菌剂
细菌蛋白
肽
序列(生物学)
生物信息学
数据科学
抗生素耐药性
人工智能
生成语法
代谢工程
背景(考古学)
系统生物学
转化研究
蛋白质工程
机器学习
镧系元素
作者
Xu Yinbiao,Jingbo Xu,ZHU Wanting,Wenhao Zhang,Peizhao Yang,Rui Chen,Xuanzheng Li,Haosheng Lin,Junhao Luo,Yongjie Ma,Kangjia Jiang,Jinyuan Hu
标识
DOI:10.1093/femsre/fuag060
摘要
The rapid emergence of antimicrobial resistance has intensified the need for alternative anti-infective strategies. Antimicrobial peptides (AMPs) offer diverse mechanisms of action, but their translation remains constrained by poor physiological stability, host toxicity, and manufacturing costs. This review examines AMP development within a design-build-test-learn (DBTL) framework that connects computational discovery, peptide production, experimental evaluation, and iterative optimization. Proteome mining and generative artificial intelligence expand the accessible sequence space, while solid-phase synthesis, engineered microbial hosts, and cell-free systems convert selected sequences into testable molecules. Chemical modification and formulation provide complementary routes to improve stability and delivery. Standardized screening and in vivo studies then establish the evidence needed to prioritize candidates and inform subsequent design and production decisions. Throughout this cycle, activity must be optimized together with toxicity, stability, manufacturability, and delivery rather than treated as the sole measure of success. This review discusses applications in medicine, agriculture, food preservation, and biotechnology, and identifies remaining limitations in data quality, assay comparability, scale-up, and translational support. The DBTL framework provides a basis for coordinating these activities, although its practical value depends on experimentally validated feedback and application-specific development criteria.
科研通智能强力驱动
Strongly Powered by AbleSci AI