{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T18:59:03Z","timestamp":1791485943602,"version":"4.3.4"},"update-to":[{"DOI":"10.1371\/journal.pcbi.1014849","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T00:00:00Z","timestamp":1791417600000}}],"reference-count":45,"publisher":"Public Library of Science (PLoS)","issue":"10","license":[{"start":{"date-parts":[[2026,10,5]],"date-time":"2026-10-05T00:00:00Z","timestamp":1791158400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:sec id=\"sec001\">\n                    <jats:title>Motivation<\/jats:title>\n                    <jats:p>Combination therapies have demonstrated improved efficacy across diverse clinical settings. Predicting antibacterial drug synergy remains difficult due to strain variability and the limited scale of experimentally tested combinations. Existing machine-learning approaches for synergy prediction often rely on permissive cross-validation schemes that allow drug pairs to re-appear across folds, leading to inflated performance metrics. A rigorous evaluation framework and scalable feature representation are needed for robust generalization.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec002\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      We assembled a curated dataset of 3,160 drug\u2013pair\u2013strain interactions covering 97 antibacterial compounds and 10 bacterial strains. We then developed\n                      <jats:bold>HALO<\/jats:bold>\n                      (\n                      <jats:italic>Held-out Antibacterial interaction Learning from latent bioactivity Observations<\/jats:italic>\n                      ), a synergy-prediction framework in which each drug pair is encoded using multi-level Chemical Checker (CC) similarity features spanning chemical, targets, networks, cellular, and clinical bioactivity domains. Using strictly nested, drug-pair heldout cross validation with fold-internal feature selection, HALO achieved consistent generalization to unseen combinations (ROC\u2013AUC\u2009=\u20090.78). Performance dropped notably under increasingly stringent heldout schemes and inflated under random data splits, underscoring the importance of rigorous evaluation for this task. Despite these conservative conditions, HALO transferred to two independent datasets, achieving ROC\u2013AUC\u2009=\u20090.90 and 0.71 and average precision\u2009=\u20090.76 and 0.94 for distinguishing synergy from antagonism. Together, these results demonstrate that multi-level bioactivity signatures provide a scalable, interpretable basis for predicting antibacterial relationships, while clarifying the realistic performance limits of current models under truly leakage-free evaluation.\n                    <\/jats:p>\n                  <\/jats:sec>","DOI":"10.1371\/journal.pcbi.1014849","type":"journal-article","created":{"date-parts":[[2026,10,5]],"date-time":"2026-10-05T17:50:09Z","timestamp":1791222609000},"page":"e1014849","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Bioactivity-driven prediction of antibacterial synergy using machine learning models"],"prefix":"10.1371","volume":"22","author":[{"given":"Hannie","family":"Yousefabadi","sequence":"first","affiliation":[{"name":"University of Tehran","place":["Tehran, Iran"],"department":["Department of Molecular Biotechnology, School of Biotechnology, College of Science"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6681-0810","authenticated-orcid":true,"given":"Mahya","family":"Mehrmohamadi","sequence":"additional","affiliation":[{"name":"University of Tehran","place":["Tehran, Iran"],"department":["Department of Molecular Biotechnology, School of Biotechnology, College of Science"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,10,5]]},"reference":[{"issue":"13","key":"pcbi.1014849.ref001","doi-asserted-by":"crossref","first-page":"1946","DOI":"10.3390\/healthcare11131946","article-title":"Antimicrobial resistance: A growing serious threat for global public health","volume":"11","author":"MDA Salam","year":"2023","journal-title":"Healthcare"},{"key":"pcbi.1014849.ref002","author":"World Health Organization","year":"2023"},{"issue":"3","key":"pcbi.1014849.ref003","doi-asserted-by":"crossref","DOI":"10.1016\/j.lanmic.2024.100992","article-title":"Addressing urgent priorities in antibiotic development: insights from WHO 2023 antibacterial clinical pipeline analyses","volume":"6","author":"D Melchiorri","year":"2025","journal-title":"Lancet Microbe"},{"issue":"1","key":"pcbi.1014849.ref004","doi-asserted-by":"crossref","first-page":"29","DOI":"10.1038\/s44259-024-00047-2","article-title":"Drug combinations targeting antibiotic resistance","volume":"2","author":"B Bogn\u00e1r","year":"2024","journal-title":"NPJ Antimicrob Resist"},{"issue":"6","key":"pcbi.1014849.ref005","doi-asserted-by":"crossref","first-page":"1391","DOI":"10.1016\/S0025-7125(05)70294-7","article-title":"Antibiotic synergy and antagonism","volume":"84","author":"JF Acar","year":"2000","journal-title":"Med Clin North Am"},{"key":"pcbi.1014849.ref006","doi-asserted-by":"crossref","first-page":"4","DOI":"10.1016\/0002-9343(83)90088-8","article-title":"Rationale for use of antimicrobial combinations","volume":"75","author":"RC Moellering","year":"1983","journal-title":"Am J Med"},{"issue":"2","key":"pcbi.1014849.ref007","doi-asserted-by":"crossref","first-page":"139","DOI":"10.1128\/CMR.1.2.139","article-title":"Antibiotic combinations: should they be tested?","volume":"1","author":"GM Eliopoulos","year":"1988","journal-title":"Clin Microbiol Rev"},{"key":"pcbi.1014849.ref008","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1016\/j.mib.2015.05.008","article-title":"Antimicrobial interactions: mechanisms and implications for drug discovery and resistance evolution","volume":"27","author":"T Bollenbach","year":"2015","journal-title":"Curr Opin Microbiol"},{"issue":"7713","key":"pcbi.1014849.ref009","doi-asserted-by":"crossref","first-page":"259","DOI":"10.1038\/s41586-018-0278-9","article-title":"Species-specific activity of antibacterial drug combinations","volume":"559","author":"AR Brochado","year":"2018","journal-title":"Nature"},{"issue":"11","key":"pcbi.1014849.ref010","doi-asserted-by":"crossref","first-page":"2196","DOI":"10.1038\/s41564-023-01486-9","article-title":"Systematic analysis of drug combinations against Gram-positive bacteria","volume":"8","author":"E Cacace","year":"2023","journal-title":"Nat Microbiol"},{"issue":"3","key":"pcbi.1014849.ref011","article-title":"Within-species variability of antibiotic interactions in Gram-negative bacteria","volume":"15","author":"P-C Tang","year":"2024","journal-title":"mBio"},{"issue":"6","key":"pcbi.1014849.ref012","article-title":"Machine learning approaches for drug combination therapies","volume":"22","author":"BG Paltun","year":"2021","journal-title":"Brief Bioinform"},{"issue":"1","key":"pcbi.1014849.ref013","doi-asserted-by":"crossref","first-page":"2428","DOI":"10.1038\/s41598-024-52814-w","article-title":"Harnessing machine learning to find synergistic combinations for FDA-approved cancer drugs","volume":"14","author":"T Abd El-Hafeez","year":"2024","journal-title":"Sci Rep"},{"issue":"3","key":"pcbi.1014849.ref014","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pcbi.1010200","article-title":"A systematic evaluation of deep learning methods for the prediction of drug synergy in cancer","volume":"19","author":"D Baptista","year":"2023","journal-title":"PLoS Comput Biol"},{"issue":"6","key":"pcbi.1014849.ref015","doi-asserted-by":"crossref","first-page":"1639","DOI":"10.1016\/j.drudis.2022.04.006","article-title":"Machine learning to design antimicrobial combination therapies: Promises and pitfalls","volume":"27","author":"JM Cantrell","year":"2022","journal-title":"Drug Discov Today"},{"issue":"6","key":"pcbi.1014849.ref016","first-page":"4","article-title":"Pharmacokinetic and molecular docking studies to design antimalarial compounds targeting Actin I","volume":"15","author":"V Guleria","year":"2021","journal-title":"Int J Health Sci (Qassim)"},{"key":"pcbi.1014849.ref017","doi-asserted-by":"crossref","first-page":"138580","DOI":"10.1016\/j.ijbiomac.2024.138580","article-title":"MDbDMRP: A novel molecular descriptor-based computational model to identify drug-miRNA relationships","volume":"287","author":"A Daroch","year":"2025","journal-title":"Int J Biol Macromol"},{"issue":"9","key":"pcbi.1014849.ref018","doi-asserted-by":"crossref","first-page":"3902","DOI":"10.1021\/acs.jmedchem.7b00204","article-title":"Prediction of antibiotic interactions using descriptors derived from molecular structure","volume":"60","author":"DJ Mason","year":"2017","journal-title":"J Med Chem"},{"issue":"5","key":"pcbi.1014849.ref019","doi-asserted-by":"crossref","first-page":"872","DOI":"10.15252\/msb.20156777","article-title":"Chemogenomics and orthology-based design of antibiotic combination therapies","volume":"12","author":"S Chandrasekaran","year":"2016","journal-title":"Mol Syst Biol"},{"key":"pcbi.1014849.ref020","doi-asserted-by":"crossref","first-page":"849006","DOI":"10.3389\/fphar.2022.849006","article-title":"Prediction of synergistic antibiotic combinations by graph learning","volume":"13","author":"J Lv","year":"2022","journal-title":"Front Pharmacol"},{"key":"pcbi.1014849.ref021","author":"MM Reuter","year":"2024"},{"issue":"9","key":"pcbi.1014849.ref022","doi-asserted-by":"crossref","first-page":"1087","DOI":"10.1038\/s41587-020-0502-7","article-title":"Extending the small-molecule similarity principle to all levels of biology with the Chemical Checker","volume":"38","author":"M Duran-Frigola","year":"2020","journal-title":"Nat Biotechnol"},{"issue":"1","key":"pcbi.1014849.ref023","doi-asserted-by":"crossref","first-page":"56","DOI":"10.1186\/s13321-023-00723-x","article-title":"Merging bioactivity predictions from cell morphology and chemical fingerprint models using similarity to training data","volume":"15","author":"S Seal","year":"2023","journal-title":"J Cheminform"},{"issue":"1","key":"pcbi.1014849.ref024","article-title":"CCSynergy: an integrative deep-learning framework enabling context-aware prediction of anti-cancer drug synergy","volume":"24","author":"S-R Hosseini","year":"2023","journal-title":"Brief Bioinform"},{"key":"pcbi.1014849.ref025","article-title":"Bio-Mol: Pretraining multimodality bioactivity profile for enhancing small molecule property prediction","author":"HF Yip","year":"2023","journal-title":"bioRxiv"},{"key":"pcbi.1014849.ref026","author":"GH Li","year":"2023"},{"issue":"1","key":"pcbi.1014849.ref027","doi-asserted-by":"crossref","first-page":"13","DOI":"10.1186\/s13321-025-00961-1","article-title":"APBIO: bioactive profiling of air pollutants through inferred bioactivity signatures and prediction of novel target interactions","volume":"17","author":"E Viesi","year":"2025","journal-title":"J Cheminform"},{"key":"pcbi.1014849.ref028","first-page":"723780","article-title":"Prediction of effective drug combinations by chemical interaction, protein interaction and target enrichment of KEGG pathways","volume":"2013","author":"L Chen","year":"2013","journal-title":"Biomed Res Int"},{"key":"pcbi.1014849.ref029","doi-asserted-by":"crossref","first-page":"869983","DOI":"10.3389\/fphar.2022.869983","article-title":"ACDB: An antibiotic combination database","volume":"13","author":"J Lv","year":"2022","journal-title":"Front Pharmacol"},{"key":"pcbi.1014849.ref030","doi-asserted-by":"crossref","DOI":"10.1093\/nar\/gkac382","article-title":"SynergyFinder 3.0: An interactive analysis and consensus interpretation of multi-drug synergies across multiple samples","volume":"50","author":"A Ianevski","year":"2022","journal-title":"Nucleic Acids Res"},{"key":"pcbi.1014849.ref031","doi-asserted-by":"crossref","first-page":"544","DOI":"10.1038\/msb.2011.71","article-title":"Systematic exploration of synergistic drug pairs","volume":"7","author":"M Cokol","year":"2011","journal-title":"Mol Syst Biol"},{"issue":"1","key":"pcbi.1014849.ref032","doi-asserted-by":"crossref","first-page":"11876","DOI":"10.1038\/s41598-019-48410-y","article-title":"Design of high-order antibiotic combinations against M. tuberculosis by ranking and exclusion","volume":"9","author":"K Yilancioglu","year":"2019","journal-title":"Sci Rep"},{"issue":"2","key":"pcbi.1014849.ref033","doi-asserted-by":"crossref","first-page":"361","DOI":"10.1111\/j.1574-6976.2007.00095.x","article-title":"Penicillin-binding proteins and beta-lactam resistance","volume":"32","author":"A Zapun","year":"2008","journal-title":"FEMS Microbiol Rev"},{"issue":"4","key":"pcbi.1014849.ref034","doi-asserted-by":"crossref","DOI":"10.1128\/spectrum.00621-22","article-title":"Penicillin plus Ceftriaxone versus Ampicillin plus ceftriaxone synergistic potential against clinical enterococcus faecalis blood isolates","volume":"10","author":"JA Cusumano","year":"2022","journal-title":"Microbiol Spectr"},{"issue":"1","key":"pcbi.1014849.ref035","doi-asserted-by":"crossref","first-page":"60","DOI":"10.1186\/s12941-024-00720-4","article-title":"Potent synergy and sustained bactericidal activity of polymyxins combined with Gram-positive only class of antibiotics versus four Gram-negative bacteria","volume":"23","author":"Y Wang","year":"2024","journal-title":"Ann Clin Microbiol Antimicrob"},{"issue":"12","key":"pcbi.1014849.ref036","doi-asserted-by":"crossref","first-page":"3793","DOI":"10.1021\/bi6025165","article-title":"Inhibition of lipid A biosynthesis as the primary mechanism of CHIR-090 antibiotic activity in Escherichia coli","volume":"46","author":"AW Barb","year":"2007","journal-title":"Biochemistry"},{"key":"pcbi.1014849.ref037","first-page":"26","article-title":"Intra-cellular accumulation and trans-epithelial transport of aivlosin, tylosin and tilmicosin","volume":"60","author":"AD Stuart","year":"2007","journal-title":"Pig J"},{"issue":"7","key":"pcbi.1014849.ref038","doi-asserted-by":"crossref","first-page":"1635","DOI":"10.1021\/acs.accounts.1c00020","article-title":"Discovery of macrolide antibiotics effective against multi-drug resistant gram-negative pathogens","volume":"54","author":"AG Myers","year":"2021","journal-title":"Acc Chem Res"},{"issue":"4","key":"pcbi.1014849.ref039","doi-asserted-by":"crossref","first-page":"593","DOI":"10.1128\/MMBR.67.4.593-656.2003","article-title":"Molecular basis of bacterial outer membrane permeability revisited","volume":"67","author":"H Nikaido","year":"2003","journal-title":"Microbiol Mol Biol Rev"},{"issue":"47","key":"pcbi.1014849.ref040","doi-asserted-by":"crossref","first-page":"18433","DOI":"10.1073\/pnas.0709412104","article-title":"Structure of the deacetylase LpxC bound to the antibiotic CHIR-090: Time-dependent inhibition and specificity in ligand binding","volume":"104","author":"AW Barb","year":"2007","journal-title":"Proc Natl Acad Sci U S A"},{"issue":"7","key":"pcbi.1014849.ref041","doi-asserted-by":"crossref","DOI":"10.1101\/cshperspect.a025304","article-title":"Antibacterial drug discovery targeting the lipopolysaccharide biosynthetic enzyme LpxC","volume":"6","author":"AL Erwin","year":"2016","journal-title":"Cold Spring Harbor Perspectives in Medicine"},{"issue":"3","key":"pcbi.1014849.ref042","doi-asserted-by":"crossref","DOI":"10.1128\/spectrum.00828-22","article-title":"TP0586532, a novel non-hydroxamate LpxC inhibitor: Potentiating effect on in vitro activity of meropenem against carbapenem-resistant Enterobacteriaceae","volume":"10","author":"I Yoshida","year":"2022","journal-title":"Microbiol Spectr"},{"issue":"3","key":"pcbi.1014849.ref043","doi-asserted-by":"crossref","first-page":"263","DOI":"10.1093\/jac\/27.3.263","article-title":"Effect of non-beta-lactam antibiotics on penicillin-binding protein synthesis of Enterococcus hirae ATCC 9790","volume":"27","author":"A Grossato","year":"1991","journal-title":"J Antimicrob Chemother"},{"issue":"11","key":"pcbi.1014849.ref044","doi-asserted-by":"crossref","first-page":"3270","DOI":"10.1038\/s41596-025-01167-3","article-title":"Integration of diverse bioactivity data into the Chemical Checker compound universe","volume":"20","author":"A Comajuncosa-Creus","year":"2025","journal-title":"Nat Protoc"},{"key":"pcbi.1014849.ref045","article-title":"A unified approach to interpreting model predictions","volume":"30","author":"SM Lundberg","year":"2017","journal-title":"Advances in neural information processing systems"}],"updated-by":[{"DOI":"10.1371\/journal.pcbi.1014849","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T00:00:00Z","timestamp":1791417600000}}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1014849","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T17:57:04Z","timestamp":1791482224000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1014849"}},"subtitle":[],"editor":[{"given":"Rahul","family":"Singh","sequence":"first","affiliation":[{"name":"CSIR-IHBT: Institute of Himalayan Bioresource Technology CSIR","place":["INDIA"]}],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,10,5]]},"references-count":45,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2026,10,5]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1014849","relation":{},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,5]]}}}