{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,26]],"date-time":"2026-08-26T10:23:11Z","timestamp":1787739791113,"version":"build-2784847793"},"reference-count":58,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,26]],"date-time":"2026-08-26T00:00:00Z","timestamp":1787702400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Pharmacol."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Cell-penetrating peptides (CPPs) are promising delivery vectors for transporting therapeutic agents across cellular membranes. However, their rational design remains challenging because the relationship between peptide sequence and translocation efficiency is highly complex and nonlinear.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>\n                      In this study, we developed an artificial intelligence-driven framework that integrates biochemical rules derived from large language models (LLMs) with conventional peptide descriptors for CPP prediction and design. Interpretable rules extracted from GPT-4o and DeepSeek were encoded as binary feature vectors and combined with sequence-based descriptors to construct hybrid machine learning models. Model performance was evaluated on the benchmark CPP924 dataset using repeated stratified cross-validation, and the optimized models were further used for\n                      <jats:italic>de novo<\/jats:italic>\n                      CPP generation. The resulting candidates were subsequently assessed using multiple established computational benchmarks.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      The top-performing hybrid classifier achieved a cross-validated accuracy of 0.91 \u00b1 0.03 (best single held-out split, 0.94) on the CPP924 dataset. The LLM-derived rules outperformed conventional physicochemical and fingerprint descriptors and matched amino-acid composition; integrating the rule and composition features yielded the best overall classifier. Using the optimized RF-GPT-Fre and RF-DS-Fre models, we generated six\n                      <jats:italic>de novo<\/jats:italic>\n                      CPP candidates that are sequence-novel (\u226453% identity to any training peptide) and retain CPP-like composition and structural features. In-silico evaluation across established tools supports computational prioritisation of these candidates for experimental testing.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>These findings demonstrate that combining LLM-derived biochemical knowledge with machine learning improves interpretable CPP prediction and candidate prioritisation. This study provides a reproducible computational strategy for peptide engineering and establishes a basis for the experimental evaluation of next-generation drug-delivery vehicles.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fphar.2026.1845913","type":"journal-article","created":{"date-parts":[[2026,8,26]],"date-time":"2026-08-26T09:54:50Z","timestamp":1787738090000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Artificial intelligence-driven prediction and design of cell-penetrating peptides for advanced drug delivery system"],"prefix":"10.3389","volume":"17","author":[{"given":"Bo","family":"Yu","sequence":"first","affiliation":[{"name":"Rheumatology and Immunology Department, Affiliated Hospital, Southwest Medical University","place":["Luzhou, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yue","family":"Lu","sequence":"additional","affiliation":[{"name":"Sichuan Treatment Center for Gynaecologic and Breast Diseases (Breast Surgery), Affiliated Hospital, Southwest Medical University","place":["Luzhou, Sichuan, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zeying","family":"Kuang","sequence":"additional","affiliation":[{"name":"Sichuan Treatment Center for Gynaecologic and Breast Diseases (Breast Surgery), Affiliated Hospital, Southwest Medical University","place":["Luzhou, Sichuan, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mengfan","family":"Zhao","sequence":"additional","affiliation":[{"name":"Sichuan Treatment Center for Gynaecologic and Breast Diseases (Breast Surgery), Affiliated Hospital, Southwest Medical University","place":["Luzhou, Sichuan, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yi","family":"Quan","sequence":"additional","affiliation":[{"name":"Sichuan Treatment Center for Gynaecologic and Breast Diseases (Breast Surgery), Affiliated Hospital, Southwest Medical University","place":["Luzhou, Sichuan, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hongyun","family":"Huang","sequence":"additional","affiliation":[{"name":"Sichuan Treatment Center for Gynaecologic and Breast Diseases (Breast Surgery), Affiliated Hospital, Southwest Medical University","place":["Luzhou, Sichuan, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,26]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"433","DOI":"10.1002\/wics.101","article-title":"Principal component analysis","volume":"2","author":"Abdi","year":"2010","journal-title":"Wiley Interdisciplinary Reviews Computational Statistics"},{"key":"B2","doi-asserted-by":"publisher","first-page":"493","DOI":"10.1038\/s41586-024-07487-w","article-title":"Accurate structure prediction of biomolecular interactions with AlphaFold 3","volume":"630","author":"Abramson","year":"2024","journal-title":"Nature"},{"key":"B3","volume-title":"Gpt-4 Technical Report, Arxiv Preprint arXiv:2303.08774","author":"Achiam","year":"2023"},{"key":"B4","doi-asserted-by":"publisher","first-page":"D1098","DOI":"10.1093\/nar\/gkv1266","article-title":"CPPsite 2.0: a repository of experimentally validated cell-penetrating peptides","volume":"44","author":"Agrawal","year":"2016","journal-title":"Nucleic Acids Research"},{"key":"B5","doi-asserted-by":"publisher","first-page":"2749","DOI":"10.1109\/TCBB.2021.3102133","article-title":"DeepCPPred: a deep learning framework for the discrimination of cell-penetrating peptides and their uptake efficiencies","volume":"19","author":"Arif","year":"2021","journal-title":"IEEE\/ACM Trans. Comput. Biol. Bioinforma."},{"key":"B6","doi-asserted-by":"publisher","first-page":"148570","DOI":"10.1109\/access.2020.3015792","article-title":"Prediction of therapeutic peptides using machine learning: computational models, datasets, and feature encodings","volume":"8","author":"Attique","year":"2020","journal-title":"Ieee Access"},{"key":"B7","doi-asserted-by":"publisher","first-page":"4363","DOI":"10.1021\/mp500439c","article-title":"Mechanisms of nanoparticle internalization and transport across an intestinal epithelial cell model: effect of size and surface charge","volume":"11","author":"Bannunah","year":"2014","journal-title":"Mol. Pharmaceutics"},{"key":"B8","doi-asserted-by":"publisher","first-page":"1209","DOI":"10.3390\/medicina57111209","article-title":"Peptide-based drug delivery systems","volume":"57","author":"Berillo","year":"2021","journal-title":"Medicina"},{"key":"B9","doi-asserted-by":"publisher","first-page":"10","DOI":"10.3390\/bdcc7010010","article-title":"Artificial intelligence in pharmaceutical and healthcare research","volume":"7","author":"Bhattamisra","year":"2023","journal-title":"Big Data Cognitive Comput."},{"key":"B10","doi-asserted-by":"publisher","first-page":"1145","DOI":"10.1016\/s0031-3203(96)00142-2","article-title":"The use of the area under the ROC curve in the evaluation of machine learning algorithms","volume":"30","author":"Bradley","year":"1997","journal-title":"Pattern Recognition"},{"key":"B11","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1023\/a:1010933404324","article-title":"Random forests","volume":"45","author":"Breiman","year":"2001","journal-title":"Mach. Learning"},{"key":"B12","doi-asserted-by":"crossref","first-page":"3121","DOI":"10.1109\/ICPR.2010.764","article-title":"The balanced accuracy and its posterior distribution","volume-title":"2010 20th International Conference on Pattern Recognition","author":"Brodersen","year":"2010"},{"key":"B13","doi-asserted-by":"publisher","first-page":"559","DOI":"10.1016\/j.addr.2004.12.001","article-title":"Tat peptide-mediated cellular delivery: back to basics","volume":"57","author":"Brooks","year":"2005","journal-title":"Adv. Drug Delivery Reviews"},{"key":"B14","doi-asserted-by":"publisher","first-page":"1","DOI":"10.2165\/11318560-000000000-00000","article-title":"Prediction of immunogenicity of therapeutic proteins: validity of computational tools","volume":"24","author":"Bryson","year":"2010","journal-title":"BioDrugs"},{"key":"B15","doi-asserted-by":"publisher","first-page":"12","DOI":"10.1002\/(sici)1097-4571(199401)45:1<12::aid-asi2>3.0.co;2-l","article-title":"The relationship between recall and precision","volume":"45","author":"Buckland","year":"1994","journal-title":"J. Am. Society Information Science"},{"key":"B16","doi-asserted-by":"publisher","first-page":"785","DOI":"10.1145\/2939672.2939785","article-title":"Xgboost: a scalable tree boosting system","author":"Chen","year":"2016","journal-title":"Proc. 22nd ACM SIGKDD, 2"},{"key":"B17","doi-asserted-by":"publisher","first-page":"e2105506","DOI":"10.1002\/advs.202105506","article-title":"Integrated design of a membrane-lytic peptide-based intravenous nanotherapeutic suppresses triple-negative breast cancer","volume":"9","author":"Chen","year":"2022","journal-title":"Adv. Sci. (Weinh)"},{"key":"B18","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s12864-019-6413-7","article-title":"The advantages of the matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation","volume":"21","author":"Chicco","year":"2020","journal-title":"BMC Genomics"},{"key":"B19","doi-asserted-by":"publisher","first-page":"239","DOI":"10.1023\/A:1024995202135","article-title":"The differential effects of cyclophosphamide, epirubicin and 5-fluorouracil on apoptotic marker (CPP-32), pro-apoptotic protein (p21 WAF\u2212 1) and anti-apoptotic protein (bcl-2) in breast cancer cells","volume":"80","author":"Chow","year":"2003","journal-title":"Breast Cancer Research Treatment"},{"key":"B20","doi-asserted-by":"publisher","first-page":"1480","DOI":"10.1039\/d0cs00556h","article-title":"Peptides as a platform for targeted therapeutics for cancer: Peptide\u2013Drug conjugates (PDCs)","volume":"50","author":"Cooper","year":"2021","journal-title":"Chem. Society Reviews"},{"key":"B21","doi-asserted-by":"publisher","first-page":"1972","DOI":"10.1021\/nn4057269","article-title":"Cell-penetrating peptides: design, synthesis, and applications","volume":"8","author":"Copolovici","year":"2014","journal-title":"ACS Nano"},{"key":"B22","first-page":"332","article-title":"Prediction of immunogenicity for therapeutic proteins: state of the art","volume":"10","author":"De Groot","year":"2007","journal-title":"Curr. Opin. Drug Discov. Dev."},{"key":"B23","doi-asserted-by":"publisher","first-page":"84","DOI":"10.1016\/s0962-8924(98)80017-2","article-title":"Trojan peptides: the penetratin system for intracellular delivery","volume":"8","author":"Derossi","year":"1998","journal-title":"Trends Cell Biology"},{"key":"B24","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s00894-014-2278-5","article-title":"AllerTOP V. 2\u2014a server for in silico prediction of allergens","volume":"20","author":"Dimitrov","year":"2014","journal-title":"J. Molecular Modeling"},{"key":"B25","doi-asserted-by":"publisher","first-page":"19","DOI":"10.1017\/s1481803500013336","article-title":"Understanding receiver operating characteristic (ROC) curves","volume":"8","author":"Fan","year":"2006","journal-title":"Can. J. Emerg. Med."},{"key":"B26","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/1479-5876-11-74","article-title":"In silico approaches for designing highly effective cell penetrating peptides","volume":"11","author":"Gautam","year":"2013","journal-title":"J. Translational Medicine"},{"key":"B27","doi-asserted-by":"publisher","first-page":"59","DOI":"10.1007\/978-1-4939-2806-4_4","article-title":"Computer-aided virtual screening and designing of cell-penetrating peptides, cell-penetrating peptides","volume":"1324","author":"Gautam","year":"2015","journal-title":"Methods Protoc."},{"key":"B28","doi-asserted-by":"publisher","first-page":"14","DOI":"10.1002\/pro.3235","article-title":"UCSF ChimeraX: meeting modern challenges in visualization and analysis","volume":"27","author":"Goddard","year":"2018","journal-title":"Protein Science"},{"key":"B29","doi-asserted-by":"publisher","first-page":"406","DOI":"10.1016\/j.tips.2017.01.003","article-title":"Cell-penetrating peptides: from basic research to clinics","volume":"38","author":"Guidotti","year":"2017","journal-title":"Trends Pharmacological Sciences"},{"key":"B30","doi-asserted-by":"publisher","first-page":"e73957","DOI":"10.1371\/journal.pone.0073957","article-title":"In silico approach for predicting toxicity of peptides and proteins","volume":"8","author":"Gupta","year":"2013","journal-title":"PloS One"},{"key":"B31","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1109\/5254.708428","article-title":"Support vector machines","volume":"13","author":"Hearst","year":"1998","journal-title":"IEEE Intelligent Syst. Their Applications"},{"key":"B32","doi-asserted-by":"crossref","DOI":"10.1002\/9781118548387","volume-title":"Applied Logistic Regression","author":"Hosmer","year":"2013"},{"key":"B33","doi-asserted-by":"publisher","first-page":"495","DOI":"10.1111\/bph.17388","article-title":"GraphCPP: the new state\u2010of\u2010the\u2010art method for cell\u2010penetrating peptide prediction via graph neural networks","volume":"182","author":"Imre","year":"2025","journal-title":"Br. J. Pharmacol."},{"key":"B34","doi-asserted-by":"publisher","first-page":"255","DOI":"10.1126\/science.aaa8415","article-title":"Machine learning: trends, perspectives, and prospects","volume":"349","author":"Jordan","year":"2015","journal-title":"Science"},{"key":"B35","doi-asserted-by":"publisher","first-page":"583","DOI":"10.1038\/s41586-021-03819-2","article-title":"Highly accurate protein structure prediction with AlphaFold","volume":"596","author":"Jumper","year":"2021","journal-title":"Nature"},{"key":"B36","doi-asserted-by":"publisher","first-page":"eadl2528","DOI":"10.1126\/science.adl2528","article-title":"Generalized biomolecular modeling and design with RoseTTAFold all-atom","volume":"384","author":"Krishna","year":"2024","journal-title":"Science"},{"key":"B37","doi-asserted-by":"publisher","first-page":"141","DOI":"10.2174\/156720107780362339","article-title":"Recent advances in protein and peptide drug delivery systems","volume":"4","author":"Kumar Malik","year":"2007","journal-title":"Curr. Drug Delivery"},{"key":"B38","first-page":"4","article-title":"Rdkit documentation","volume":"1","author":"Landrum","year":"2013","journal-title":"Release"},{"key":"B39","volume-title":"Deepseek-v3 Technical Report, Arxiv Preprint arXiv:2412.19437","author":"Liu","year":""},{"key":"B40","doi-asserted-by":"publisher","first-page":"137069","DOI":"10.1016\/j.ijbiomac.2024.137069","article-title":"GPT4Kinase: high-Accuracy prediction of inhibitor-kinase binding affinity utilizing large language model","volume":"282","author":"Liu","year":"","journal-title":"Int. J. Biol. Macromol."},{"key":"B41","volume-title":"Umap: Uniform Manifold Approximation and Projection for Dimension Reduction","author":"McInnes","year":"2018"},{"key":"B42","doi-asserted-by":"publisher","first-page":"850","DOI":"10.1016\/j.drudis.2012.03.002","article-title":"Cell-penetrating peptides: classes, origin, and current landscape","volume":"17","author":"Milletti","year":"2012","journal-title":"Drug Discovery Today"},{"key":"B43","doi-asserted-by":"publisher","first-page":"70","DOI":"10.1002\/pro.3943","article-title":"UCSF ChimeraX: structure visualization for researchers, educators, and developers","volume":"30","author":"Pettersen","year":"2021","journal-title":"Protein Science"},{"key":"B44","doi-asserted-by":"publisher","first-page":"175","DOI":"10.7785\/tcrt.2012.500247","article-title":"HER2+ breast cancer therapy: by CPP-ZFN mediated targeting of mTOR?","volume":"11","author":"Puria","year":"2012","journal-title":"Technol. Cancer Research and Treatment"},{"key":"B45","doi-asserted-by":"publisher","first-page":"1818","DOI":"10.2174\/1386207325666211207153943","article-title":"Artificial intelligence (AI) in drugs and pharmaceuticals","volume":"25","author":"Sahu","year":"2022","journal-title":"Comb. Chemistry and High Throughput Screening"},{"key":"B46","doi-asserted-by":"publisher","first-page":"327","DOI":"10.1016\/j.jclinepi.2005.09.005","article-title":"\u201cPrecision\u201d and \u201caccuracy\u201d: two terms that are neither","volume":"59","author":"Streiner","year":"2006","journal-title":"J. Clinical Epidemiology"},{"key":"B47","doi-asserted-by":"publisher","first-page":"150","DOI":"10.1016\/j.bbrc.2016.06.035","article-title":"Prediction of cell-penetrating peptides with feature selection techniques","volume":"477","author":"Tang","year":"2016","journal-title":"Biochem. Biophysical Research Communications"},{"key":"B48","doi-asserted-by":"publisher","first-page":"351","DOI":"10.3390\/molecules24020351","article-title":"Peptide-based drug-delivery systems in biotechnological applications: recent advances and perspectives","volume":"24","author":"Tesauro","year":"2019","journal-title":"Molecules"},{"key":"B49","doi-asserted-by":"publisher","first-page":"40","DOI":"10.3758\/bf03213026","article-title":"Theoretical analysis of an alphabetic confusion matrix","volume":"9","author":"Townsend","year":"1971","journal-title":"Percept. and Psychophys."},{"key":"B50","first-page":"2579","article-title":"Visualizing data using t-SNE","volume":"9","author":"Van der Maaten","year":"2008","journal-title":"J. Machine Learning Research"},{"key":"B51","first-page":"13","article-title":"Dimensionality reduction: a comparative review","volume":"10","author":"Van Der Maaten","year":"2009","journal-title":"J. Machine Learning Research"},{"key":"B52","doi-asserted-by":"publisher","first-page":"1916","DOI":"10.3390\/pharmaceutics15071916","article-title":"Artificial intelligence in pharmaceutical technology and drug delivery design","volume":"15","author":"Vora","year":"2023","journal-title":"Pharmaceutics"},{"key":"B53","doi-asserted-by":"publisher","first-page":"2044","DOI":"10.1021\/acs.jproteome.7b00019","article-title":"CPPred-RF: a sequence-based predictor for identifying cell-penetrating peptides and their uptake efficiency","volume":"16","author":"Wei","year":"2017","journal-title":"J. Proteome Res."},{"key":"B54","doi-asserted-by":"publisher","first-page":"217","DOI":"10.2174\/1389200219666181010114750","article-title":"The development of machine learning methods in cell-penetrating peptides identification: a brief review","volume":"20","author":"Wei","year":"2019","journal-title":"Curr. Drug Metab."},{"key":"B55","doi-asserted-by":"publisher","first-page":"697","DOI":"10.3389\/fphar.2020.00697","article-title":"Cell-penetrating peptides in diagnosis and treatment of human diseases: from preclinical research to clinical application","volume":"11","author":"Xie","year":"2020","journal-title":"Front. Pharmacology"},{"key":"B56","doi-asserted-by":"publisher","first-page":"1553853","DOI":"10.3389\/fphar.2025.1553853","article-title":"Trends in the research and development of peptide drug conjugates: artificial intelligence aided design","volume":"16","author":"Zhang","year":"2025","journal-title":"Front. Pharmacol."},{"key":"B57","first-page":"6","article-title":"Recall, precision and average precision","volume":"2","author":"Zhu","year":"2004","journal-title":"Dep. Statistics Actuar. Sci. Univ. Waterloo, Waterloo"},{"key":"B58","doi-asserted-by":"publisher","first-page":"744","DOI":"10.1007\/s12539-024-00673-4","article-title":"EnDM-CPP: a multi-view explainable framework based on deep learning and machine learning for identifying cell-penetrating peptides with transformers and analyzing sequence information","volume":"17","author":"Zhu","year":"2024","journal-title":"Interdiscip. Sci"}],"container-title":["Frontiers in Pharmacology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fphar.2026.1845913\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,26]],"date-time":"2026-08-26T09:54:53Z","timestamp":1787738093000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fphar.2026.1845913\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,26]]},"references-count":58,"alternative-id":["10.3389\/fphar.2026.1845913"],"URL":"https:\/\/doi.org\/10.3389\/fphar.2026.1845913","relation":{},"ISSN":["1663-9812"],"issn-type":[{"value":"1663-9812","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,26]]},"article-number":"1845913"}}