{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T19:26:01Z","timestamp":1792092361714,"version":"4.4.0"},"posted":{"date-parts":[[2025,9,23]]},"group-title":"Molecular Biology","reference-count":45,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2025,9,23]],"date-time":"2025-09-23T00:00:00Z","timestamp":1758585600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2025,9,23]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>\n                  Antimicrobial resistance (AMR) is projected to cause up to 10 million deaths annually by 2050, underscoring the urgent need for new antibiotics. Here we present ApexAmphion, a deep-learning framework for\n                  <jats:italic>de novo<\/jats:italic>\n                  design of antibiotics that couples a 6.4-billion-parameter protein language model with reinforcement learning. The model is first fine-tuned on curated peptide data to capture antimicrobial sequence regularities, then optimised with proximal policy optimization against a composite reward that combines predictions from a learned minimum inhibitory concentration (MIC) classifier with differentiable physicochemical objectives.\n                  <jats:italic>In vitro<\/jats:italic>\n                  evaluation of 100 designed peptides showed low MIC values (nanomolar range in some cases) for all candidates (100% hit rate). Moreover, 99 our of 100 compounds exhibited broad-spectrum antimicrobial activity against at least two clinically relevant bacteria. The lead molecules killed bacteria primarily by potently targeting the cytoplasmic membrane. By unifying generation, scoring and multi-objective optimization with deep reinforcement learning in a single pipeline, our approach rapidly produces diverse, potent candidates, offering a scalable route to peptide antibiotics and a platform for iterative steering toward potency and developability within hours.\n                <\/jats:p>","DOI":"10.1101\/2025.09.23.678086","type":"posted-content","created":{"date-parts":[[2025,9,24]],"date-time":"2025-09-24T04:40:14Z","timestamp":1758688814000},"source":"Crossref","is-referenced-by-count":8,"title":["A deep reinforcement learning platform for antibiotic discovery"],"prefix":"10.64898","author":[{"given":"Hanqun","family":"Cao","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marcelo","family":"D. T. Torres","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jingjie","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zijun","family":"Gao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7240-3915","authenticated-orcid":false,"given":"Fang","family":"Wu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chunbin","family":"Gu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jure","family":"Leskovec","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yejin","family":"Choi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cesar","family":"de la Fuente-Nunez","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Guangyong","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Pheng-Ann","family":"Heng","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2025092610050719000_2025.09.23.678086v1.1","unstructured":"O\u2019Neill, J . Tackling drug-resistant infections globally: final report and recommendations. Rev. Antimicrob. Resist. (2016)."},{"key":"2025092610050719000_2025.09.23.678086v1.2","unstructured":"World Health Organization and Food and Agriculture Organization of the United Nations and World Organisation for Animal Health. Technical Series on Antimicrobial Resistance: The Impact of Antimicrobial Resistance on Global Health and Food Security. (2019)."},{"key":"2025092610050719000_2025.09.23.678086v1.3","doi-asserted-by":"crossref","unstructured":"de la Fuente-Nunez, C. & Collins, J. J . Essay: Using Machine Learning for Antibiotic Discovery. Phys. Rev. Lett. 135, 030001 (2025).","DOI":"10.1103\/y3fg-s9vg"},{"key":"2025092610050719000_2025.09.23.678086v1.4","first-page":"854","article-title":"Deep-learning-enabled antibiotic discovery through molecular de-extinction. Nat","volume":"8","year":"2024","journal-title":"Biomed. Eng"},{"key":"2025092610050719000_2025.09.23.678086v1.5","doi-asserted-by":"publisher","DOI":"10.1126\/science.adh1114"},{"key":"2025092610050719000_2025.09.23.678086v1.6","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-018-03746-3"},{"key":"2025092610050719000_2025.09.23.678086v1.7","doi-asserted-by":"crossref","unstructured":"Mahlapuu, M. , H\u00e5kansson, J. , Ringstad, L . & Bj\u00f6rn, C. Antimicrobial peptides: an emerging category of therapeutic agents . Front. Cell. Infect. Microbiol. 6, 194 (2016).","DOI":"10.3389\/fcimb.2016.00194"},{"key":"2025092610050719000_2025.09.23.678086v1.8","doi-asserted-by":"crossref","unstructured":"Huang, J. et al. Identification of potent antimicrobial peptides via a machine-learning pipeline that mines the entire space of peptide sequences. Nat . Biomed. Eng. 1\u201314 (2023).","DOI":"10.1038\/s41551-022-00991-2"},{"key":"2025092610050719000_2025.09.23.678086v1.9","unstructured":"Wang, B. et al. Explainable deep learning and virtual evolution identifies antimicrobial peptides with activity against multidrug-resistant human pathogens. Nat. Microbiol. 1\u201316 (2025)."},{"key":"2025092610050719000_2025.09.23.678086v1.10","doi-asserted-by":"crossref","first-page":"3420","DOI":"10.1038\/s41467-025-58804-4","article-title":"Pre-trained molecular representations enable antimicrobial discovery","volume":"16","year":"2025","journal-title":"Nat. Commun"},{"key":"2025092610050719000_2025.09.23.678086v1.11","doi-asserted-by":"publisher","DOI":"10.7717\/peerj.10555"},{"key":"2025092610050719000_2025.09.23.678086v1.12","doi-asserted-by":"publisher","DOI":"10.1016\/J.CELL.2024.05.013"},{"key":"2025092610050719000_2025.09.23.678086v1.13","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2024.07.027"},{"key":"2025092610050719000_2025.09.23.678086v1.14","doi-asserted-by":"publisher","DOI":"10.1016\/j.chom.2020.11.002"},{"key":"2025092610050719000_2025.09.23.678086v1.15","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btaa917"},{"key":"2025092610050719000_2025.09.23.678086v1.16","doi-asserted-by":"crossref","unstructured":"Cesaro, A. , Wan, F. , Torres, M. D. & de la Fuente-Nunez, C . Design of multimodal antibiotics against intracellular infections using deep learning. bioRxiv 2024\u201312 (2024).","DOI":"10.1101\/2024.12.20.629780"},{"key":"2025092610050719000_2025.09.23.678086v1.17","doi-asserted-by":"crossref","first-page":"2153","DOI":"10.1038\/s41564-025-02061-0","article-title":"Deep learning reveals antibiotics in the archaeal proteome","volume":"10","year":"2025","journal-title":"Nat. Microbiol"},{"key":"2025092610050719000_2025.09.23.678086v1.18","doi-asserted-by":"crossref","first-page":"6446","DOI":"10.1038\/s41467-025-60051-6","article-title":"Computational exploration of global venoms for antimicrobial discovery with Venomics artificial intelligence","volume":"16","year":"2025","journal-title":"Nat. Commun"},{"key":"2025092610050719000_2025.09.23.678086v1.19","doi-asserted-by":"publisher","DOI":"10.1111\/j.1747-0285.2010.01044.x"},{"key":"2025092610050719000_2025.09.23.678086v1.20","doi-asserted-by":"crossref","unstructured":"Boone, K. , Wisdom, C. , Camarda, K. , Spencer, P. & Tamerler, C . Combining genetic algorithm with machine learning strategies for designing potent antimicrobial peptides. BMC Bioinformatics 22, 239 (2021).","DOI":"10.1186\/s12859-021-04156-x"},{"key":"2025092610050719000_2025.09.23.678086v1.21","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-023-36994-z"},{"key":"2025092610050719000_2025.09.23.678086v1.22","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.0c01441"},{"key":"2025092610050719000_2025.09.23.678086v1.23","unstructured":"Das, P. et al. PepCVAE: Semi-Supervised Targeted Design of Antimicrobial Peptide Sequences. Preprint at http:\/\/arxiv.org\/abs\/1810.07743 (2018)."},{"key":"2025092610050719000_2025.09.23.678086v1.24","doi-asserted-by":"crossref","unstructured":"Wang, R. et al. Diff-AMP: tailored designed antimicrobial peptide framework with all-in-one generation, identification, prediction and optimization. Brief. Bioinform. 25, bbae078 (2024).","DOI":"10.1093\/bib\/bbae078"},{"key":"2025092610050719000_2025.09.23.678086v1.25","doi-asserted-by":"crossref","first-page":"7538","DOI":"10.1038\/s41467-024-51933-2","article-title":"A foundation model identifies broad-Spectrum antimicrobial peptides against drug-resistant bacterial infection","volume":"15","year":"2024","journal-title":"Nat. Commun"},{"key":"2025092610050719000_2025.09.23.678086v1.26","doi-asserted-by":"crossref","unstructured":"Wang, J. et al. Discovery of antimicrobial peptides with notable antibacterial potency by an LLM-based foundation model. Sci. Adv. 11, eads8932 (2025).","DOI":"10.1126\/sciadv.ads8932"},{"key":"2025092610050719000_2025.09.23.678086v1.27","doi-asserted-by":"crossref","unstructured":"Chen, T. , Vure, P. , Pulugurta, R. & Chatterjee, P . AMP-diffusion: Integrating latent diffusion with protein language models for antimicrobial peptide generation. bioRxiv 2024\u201303 (2024).","DOI":"10.1101\/2024.03.03.583201"},{"key":"2025092610050719000_2025.09.23.678086v1.28","doi-asserted-by":"publisher","DOI":"10.1101\/2024.11.27.625757"},{"key":"2025092610050719000_2025.09.23.678086v1.29","doi-asserted-by":"publisher","DOI":"10.1093\/NAR\/GKAA991"},{"key":"2025092610050719000_2025.09.23.678086v1.30","doi-asserted-by":"publisher","DOI":"10.1093\/NAR\/GKAB651"},{"key":"2025092610050719000_2025.09.23.678086v1.31","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkv1278"},{"key":"2025092610050719000_2025.09.23.678086v1.32","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkp1021"},{"key":"2025092610050719000_2025.09.23.678086v1.33","doi-asserted-by":"crossref","unstructured":"Peng, S. & Rajjou, L . Unifying antimicrobial peptide datasets for robust deep learning-based classification. Data Brief 56, 110822 (2024).","DOI":"10.1016\/j.dib.2024.110822"},{"key":"2025092610050719000_2025.09.23.678086v1.34","doi-asserted-by":"crossref","first-page":"1213","DOI":"10.1038\/s41597-025-05528-1","article-title":"A comprehensive dataset of therapeutic peptides on multi-function property and structure information","volume":"12","year":"2025","journal-title":"Sci. Data"},{"key":"2025092610050719000_2025.09.23.678086v1.35","doi-asserted-by":"crossref","first-page":"968","DOI":"10.1016\/j.cels.2023.10.002","article-title":"ProGen2: exploring the boundaries of protein language models","volume":"14","year":"2023","journal-title":"Cell Syst"},{"key":"2025092610050719000_2025.09.23.678086v1.36","first-page":"27730","article-title":"Training language models to follow instructions with human feedback","volume":"35","year":"2022","journal-title":"Adv. Neural Inf. Process. Syst"},{"key":"2025092610050719000_2025.09.23.678086v1.37","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2025.07.033"},{"key":"2025092610050719000_2025.09.23.678086v1.38","doi-asserted-by":"publisher","DOI":"10.1093\/NAR\/GKAC933"},{"key":"2025092610050719000_2025.09.23.678086v1.39","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/bty179"},{"key":"2025092610050719000_2025.09.23.678086v1.40","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btx285"},{"key":"2025092610050719000_2025.09.23.678086v1.41","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btm098"},{"key":"2025092610050719000_2025.09.23.678086v1.42","doi-asserted-by":"publisher","DOI":"10.1038\/nbt.3988"},{"key":"2025092610050719000_2025.09.23.678086v1.43","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/bts141"},{"key":"2025092610050719000_2025.09.23.678086v1.44","doi-asserted-by":"crossref","unstructured":"Ye, G. et al. LAMP2: a major update of the database linking antimicrobial peptides. Database 2020, baaa061 (2020).","DOI":"10.1093\/database\/baaa061"},{"key":"2025092610050719000_2025.09.23.678086v1.45","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkab1080"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2025.09.23.678086","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T09:39:02Z","timestamp":1768469942000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.1101\/2025.09.23.678086"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9,23]]},"references-count":45,"URL":"https:\/\/doi.org\/10.1101\/2025.09.23.678086","relation":{},"subject":[],"published":{"date-parts":[[2025,9,23]]},"subtype":"preprint"}}