{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,22]],"date-time":"2026-08-22T16:25:28Z","timestamp":1787415928594,"version":"build-2736575974"},"reference-count":54,"publisher":"American Chemical Society (ACS)","issue":"8","license":[{"start":{"date-parts":[[2026,7,22]],"date-time":"2026-07-22T00:00:00Z","timestamp":1784678400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"funder":[{"DOI":"10.13039\/100013405","name":"Rijksdienst voor Ondernemend Nederland","doi-asserted-by":"publisher","award":["NA"],"award-info":[{"award-number":["NA"]}],"id":[{"id":"10.13039\/100013405","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/na","name":"dsm-firmenich","doi-asserted-by":"publisher","award":["NA"],"award-info":[{"award-number":["NA"]}],"id":[{"id":"10.13039\/na","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,8,21]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Combinatorial pathway optimization is a powerful approach in metabolic engineering to improve strain performance. While machine learning (ML) has shown promise in guiding the Design\u2013Build\u2013Test\u2013Learn (DBTL) cycle, most applications have been limited to small design spaces, thereby restricting the potential of predictive and exploration-exploitation strategies. In this work, we applied two DBTL cycles to optimize p-coumaric acid production in Saccharomyces cerevisiae. The first cycle involved constructing a large combinatorial library of 18 genes and 20 promoters (170 million possible designs). In the second cycle, we employed a gradient bandit-based machine learning recommendation strategy, tuned to balance exploration and exploitation. Our results show that this balanced strategy outperforms greedy, feature importance-based approaches, leading to greater diversity in strain performance and improved top-producer identification. Notably, applying the same strategy to an alternative parent strain yielded the highest p-coumaric acid titer (1.23 g\/L), a 2.37-fold improvement over the original. These findings highlight the value of ML-guided exploration in large design spaces and demonstrate that balancing exploration and exploitation is critical for successful strain optimization.<\/jats:p>","DOI":"10.1021\/acssynbio.5c00864","type":"journal-article","created":{"date-parts":[[2026,7,22]],"date-time":"2026-07-22T23:48:14Z","timestamp":1784764094000},"page":"3146-3157","source":"Crossref","is-referenced-by-count":0,"title":["Machine-Learning-Assisted Pathway Optimization in\nLarge Combinatorial Design Spaces: A\n                    <i>p<\/i>\n                    -Coumaric\nAcid Case Study"],"prefix":"10.1021","volume":"15","author":[{"ORCID":"https:\/\/orcid.org\/0009-0001-2887-0193","authenticated-orcid":true,"given":"Paul","family":"van Lent","sequence":"first","affiliation":[{"name":"Delft University of Technology , , ,","place":["Delft, The Netherlands, 2628 AT"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rianne","family":"van der Hoek","sequence":"additional","affiliation":[{"name":"dsm-firmenich , , ,","place":["Delft, The Netherlands, 2613 AX"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sara","family":"Moreno Paz","sequence":"additional","affiliation":[{"name":"dsm-firmenich , , ,","place":["Delft, The Netherlands, 2613 AX"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Irsan","family":"Kooi","sequence":"additional","affiliation":[{"name":"dsm-firmenich , , ,","place":["Delft, The Netherlands, 2613 AX"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Moniek","family":"Jonkers","sequence":"additional","affiliation":[{"name":"dsm-firmenich , , ,","place":["Delft, The Netherlands, 2613 AX"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Priscilla","family":"Zwartjens","sequence":"additional","affiliation":[{"name":"dsm-firmenich , , ,","place":["Delft, The Netherlands, 2613 AX"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joep","family":"Schmitz","sequence":"additional","affiliation":[{"name":"dsm-firmenich , , ,","place":["Delft, The Netherlands, 2613 AX"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7205-7431","authenticated-orcid":true,"given":"Thomas","family":"Abeel","sequence":"additional","affiliation":[{"name":"Delft University of Technology , , ,","place":["Delft, The Netherlands, 2628 AT"]},{"name":"Broad Institute of MIT and Harvard , , ,","place":["Cambridge, United States, 02142"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"316","published-online":{"date-parts":[[2026,7,22]]},"reference":[{"key":"2026082212130768300_cit1","doi-asserted-by":"publisher","first-page":"123","DOI":"10.1016\/j.ymben.2014.11.011","article-title":"Principal component\nanalysis of proteomics (PCAP) as a tool to direct metabolic engineering","volume":"28","author":"Alonso-Gutierrez","year":"2015","journal-title":"Metab. Eng."},{"key":"2026082212130768300_cit2","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-020-15798-5","article-title":"Large scale active-learning-guided\nexploration for in vitro protein production optimization","volume":"11","author":"Borkowski","year":"2020","journal-title":"Nat. Commun."},{"key":"2026082212130768300_cit3","doi-asserted-by":"crossref","unstructured":"Chen, T.; Guestrin, C.; et al.\n          Xgboost:\nA scalable tree boosting system. In  Proceedings\nof the 22nd acm sigkdd International Conference on Knowledge Discovery\nand Data Mining2016; pp 785\u2013794.","DOI":"10.1145\/2939672.2939785"},{"key":"2026082212130768300_cit4","doi-asserted-by":"publisher","DOI":"10.1093\/femsyr\/foz060","article-title":"Anaerobic growth of Saccharomyces cerevisiae CEN. PK113\u20137D does\nnot depend on synthesis or supplementation of unsaturated fatty acids","volume":"19","author":"Dekker","year":"2019","journal-title":"FEMS Yeast Res."},{"key":"2026082212130768300_cit5","doi-asserted-by":"publisher","first-page":"563","DOI":"10.1146\/annurev-biochem-062608-095938","article-title":"High-throughput\nmetabolic engineering:\nadvances in small-molecule screening and selection","volume":"79","author":"Dietrich","year":"2010","journal-title":"Annu. Rev. Biochem."},{"key":"2026082212130768300_cit6","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-50698-y","article-title":"Machine learning-guided\nco-optimization of fitness and diversity facilitates combinatorial\nlibrary design in enzyme engineering","volume":"15","author":"Ding","year":"2024","journal-title":"Nat. Commun."},{"key":"2026082212130768300_cit7","unstructured":"EUR-Lex\n- 52024DC0137 - EN - EUR-Lex. https:\/\/eur-lex.europa.eu\/legal-content\/EN\/TXT\/?uri=celex:52024DC0137."},{"key":"2026082212130768300_cit8","unstructured":"Frazier, P. I.\n          \n          A tutorial\non Bayesian optimization. arXiv:1807.02811. arXiv.org\ne-Print archive. https:\/\/arxiv.org\/abs\/1807.02811. 2018."},{"key":"2026082212130768300_cit9","doi-asserted-by":"publisher","first-page":"355","DOI":"10.1002\/yea.320110408","article-title":"Studies on the transformation\nof intact yeast cells by the LiAc\/SS-DNA\/PEG procedure","volume":"11","author":"Gietz","year":"1995","journal-title":"Yeast"},{"key":"2026082212130768300_cit10","doi-asserted-by":"publisher","first-page":"610","DOI":"10.1111\/1751-7915.12385","article-title":"Prospects of microbial\ncell factories developed through\nsystems metabolic engineering","volume":"9","author":"Gustavsson","year":"2016","journal-title":"Microb. Biotechnol."},{"key":"2026082212130768300_cit11","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-019-13189-z","article-title":"Towards\na fully automated\nalgorithm driven platform for biosystems design","volume":"10","author":"HamediRad","year":"2019","journal-title":"Nat. Commun."},{"key":"2026082212130768300_cit12","unstructured":"Head, T.; MechCoder, G. L.; Shcherbatyi, I.; et al.\n          scikit-optimize\/scikit-optimize:\nv0. 5.2\n          Version v0\n          2018; Vol. 5."},{"key":"2026082212130768300_cit13","doi-asserted-by":"publisher","first-page":"565","DOI":"10.1007\/s00449-025-03128-2","article-title":"Metabolic engineering\nof Escherichia coli for enhanced production\nof p-coumaric acid via l-phenylalanine biosynthesis\npathway","volume":"48","author":"Jeong","year":"2025","journal-title":"Bioprocess Biosyst. Eng."},{"key":"2026082212130768300_cit14","doi-asserted-by":"publisher","first-page":"142","DOI":"10.1016\/j.copbio.2017.06.014","article-title":"Combinatorial pathway\noptimization\nfor streamlined metabolic engineering","volume":"47","author":"Jeschek","year":"2017","journal-title":"Curr.\nOpin. Biotechnol."},{"key":"2026082212130768300_cit15","doi-asserted-by":"publisher","DOI":"10.1038\/ncomms11163","article-title":"Rationally reduced libraries for\ncombinatorial pathway optimization minimizing experimental effort","volume":"7","author":"Jeschek","year":"2016","journal-title":"Nat. Commun."},{"key":"2026082212130768300_cit16","doi-asserted-by":"publisher","DOI":"10.1016\/j.biortech.2022.128455","article-title":"Machine learning-guided\nprediction of potential engineering targets for microbial production\nof lycopene","volume":"369","author":"Kang","year":"2023","journal-title":"Bioresour. Technol."},{"key":"2026082212130768300_cit17","doi-asserted-by":"publisher","first-page":"1333","DOI":"10.2174\/1385272826666221012145959","article-title":"p-Coumaric\nacid: A naturally occurring chemical\nwith potential therapeutic applications","volume":"26","author":"Kaur","year":"2022","journal-title":"Curr.\nOrg. Chem."},{"key":"2026082212130768300_cit18","doi-asserted-by":"publisher","first-page":"425","DOI":"10.1016\/j.tibtech.2022.12.014","article-title":"Metabolic engineering\nfor sustainability and health","volume":"41","author":"Kim","year":"2023","journal-title":"Trends Biotechnol."},{"issue":"11","key":"2026082212130768300_cit19","doi-asserted-by":"publisher","first-page":"1339","DOI":"10.1016\/j.tibtech.2024.04.007","article-title":"TRYing to evaluate\nproduction costs in microbial biotechnology","volume":"42","author":"Konzock","year":"2024","journal-title":"Trends Biotechnol."},{"key":"2026082212130768300_cit20","doi-asserted-by":"publisher","first-page":"722","DOI":"10.1101\/gr.215087.116","article-title":"Canu:\nscalable and accurate\nlong-read assembly via adaptive k-mer weighting and repeat separation","volume":"27","author":"Koren","year":"2017","journal-title":"Genome Res."},{"key":"2026082212130768300_cit21","doi-asserted-by":"publisher","first-page":"128","DOI":"10.1016\/j.ymben.2023.03.005","article-title":"Elucidating yeast glycolytic\ndynamics at steady state growth and glucose pulses through kinetic\nmetabolic modeling","volume":"77","author":"Lao-Martil","year":"2023","journal-title":"Metab. Eng."},{"key":"2026082212130768300_cit22","doi-asserted-by":"publisher","first-page":"1061","DOI":"10.1038\/nbt.3365","article-title":"Systems strategies for developing\nindustrial microbial\nstrains","volume":"33","author":"Lee","year":"2015","journal-title":"Nat. Biotechnol."},{"key":"2026082212130768300_cit23","doi-asserted-by":"publisher","first-page":"2588","DOI":"10.1021\/acssynbio.3c00186","article-title":"Simulated design-build-test-learn\ncycles for consistent comparison of machine learning methods in metabolic\nengineering","volume":"12","author":"van Lent","year":"2023","journal-title":"ACS Synth. Biol."},{"key":"2026082212130768300_cit24","doi-asserted-by":"publisher","first-page":"390","DOI":"10.1038\/msb.2010.47","article-title":"Omic data from evolved E. coli are\nconsistent with computed optimal growth\nfrom genome-scale models","volume":"6","author":"Lewis","year":"2010","journal-title":"Mol. Syst. Biol."},{"issue":"9","key":"2026082212130768300_cit25","doi-asserted-by":"publisher","DOI":"10.1016\/j.cels.2025.101387","article-title":"Evaluation of machine learning-assisted directed evolution across\ndiverse combinatorial landscapes","volume":"16","author":"Li","year":"2024","journal-title":"Cell Syst."},{"key":"2026082212130768300_cit26","doi-asserted-by":"publisher","first-page":"18","DOI":"10.3390\/e23010018","article-title":"Explainable ai: A review of machine\nlearning interpretability methods","volume":"23","author":"Linardatos","year":"2021","journal-title":"Entropy"},{"key":"2026082212130768300_cit27","doi-asserted-by":"publisher","first-page":"417","DOI":"10.1016\/bs.mie.2018.06.009","article-title":"Modular pathway rewiring of yeast for amino acid production","volume":"608","author":"Liu","year":"2018","journal-title":"Methods Enzymol."},{"key":"2026082212130768300_cit28","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-019-12961-5","article-title":"Rewiring carbon metabolism in yeast for high\nlevel production of aromatic chemicals","volume":"10","author":"Liu","year":"2019","journal-title":"Nat.\nCommun."},{"key":"2026082212130768300_cit29","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-019-11581-3","article-title":"A consensus S. cerevisiae metabolic\nmodel Yeast8 and its ecosystem\nfor comprehensively probing cellular metabolism","volume":"10","author":"Lu","year":"2019","journal-title":"Nat. Commun."},{"key":"2026082212130768300_cit30","unstructured":"Lundberg, S. M.; Lee, S.-I.\n          A unified approach\nto interpreting model predictions\n          Adv. Neural\nInf. Process. Syst.\n          2017; Vol. 3010.48550\/arXiv.1705.07874."},{"key":"2026082212130768300_cit31","doi-asserted-by":"crossref","first-page":"1077","DOI":"10.1002\/yea.320111109","article-title":"Sequencing analysis of a 24.7 kb fragment of yeast chromosome XIV\nidentifies six known genes, a new member of the hexose transporter\nfamily and ten new open reading frames","volume":"11.11","author":"Maftahi","year":"1995","journal-title":"Yeast"},{"key":"2026082212130768300_cit32","doi-asserted-by":"publisher","first-page":"192","DOI":"10.1016\/j.ymben.2023.08.003","article-title":"Fine-tuning of p-coumaric acid synthesis to increase\n(2S)-naringenin production in yeast","volume":"79","author":"Mao","year":"2023","journal-title":"Metab. Eng."},{"key":"2026082212130768300_cit33","doi-asserted-by":"publisher","DOI":"10.1111\/1751-7915.14424","article-title":"Combinatorial optimization\nof pathway, process and media for the production of p-coumaric acid by Saccharomyces cerevisiae","volume":"17","author":"Moreno-Paz","year":"2024","journal-title":"Microb. Biotechnol."},{"key":"2026082212130768300_cit34","doi-asserted-by":"publisher","first-page":"1312","DOI":"10.1021\/acssynbio.4c00035","article-title":"Machine learning-guided\noptimization of p-coumaric acid production in yeast","volume":"13","author":"Moreno-Paz","year":"2024","journal-title":"ACS Synth. Biol."},{"key":"2026082212130768300_cit35","doi-asserted-by":"publisher","first-page":"153","DOI":"10.1007\/BF00163778","article-title":"A further characterization of Borda ranking method","volume":"36","author":"Nitzan","year":"1981","journal-title":"Public Choice"},{"key":"2026082212130768300_cit36","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1003098","article-title":"Consistent\nestimation of Gibbs energy using component contributions","volume":"9","author":"Noor","year":"2013","journal-title":"PLoS Comput. Biol."},{"key":"2026082212130768300_cit37","doi-asserted-by":"publisher","first-page":"1337","DOI":"10.1021\/acssynbio.9b00020","article-title":"Lessons from two design-build-test-learn\ncycles of dodecanol production in Escherichia coli aided by machine learning","volume":"8","author":"Opgenorth","year":"2019","journal-title":"ACS Synth. Biol."},{"key":"2026082212130768300_cit38","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-022-31245-z","article-title":"A versatile active learning\nworkflow for optimization of genetic and metabolic networks","volume":"13","author":"Pandi","year":"2022","journal-title":"Nat. Commun."},{"key":"2026082212130768300_cit39","first-page":"2825","article-title":"Scikit-learn: Machine\nLearning in Python","volume":"12","author":"Pedregosa","year":"2011","journal-title":"J.\nMach. Learn. Res."},{"key":"2026082212130768300_cit40","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-020-18008-4","article-title":"A machine learning Automated\nRecommendation Tool for synthetic biology","volume":"11","author":"Radivojevic\u0301","year":"2020","journal-title":"Nat.\nCommun."},{"key":"2026082212130768300_cit41","first-page":"7926","article-title":"Balanced\nmse for imbalanced\nvisual regression","volume-title":"Proceedings of the IEEE\/CVF\nConference on Computer Vision and Pattern Recognition","author":"Ren","year":"2022"},{"key":"2026082212130768300_cit42","doi-asserted-by":"publisher","first-page":"181","DOI":"10.1016\/j.ymben.2015.08.003","article-title":"Establishment\nof a yeast\nplatform strain for production of p-coumaric acid\nthrough metabolic engineering of aromatic amino acid biosynthesis","volume":"31","author":"Rodriguez","year":"2015","journal-title":"Metab. Eng."},{"key":"2026082212130768300_cit43","unstructured":"Roubos, J. A.; Van Peij, N. N. M. E.\n          Method\nfor achieving improved polypeptide expression. US\nPatent US8,812,247, 2014."},{"key":"2026082212130768300_cit44","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1010177","article-title":"Strain design optimization\nusing reinforcement learning","volume":"18","author":"Sabzevari","year":"2022","journal-title":"PLoS Comput. Biol."},{"key":"2026082212130768300_cit45","doi-asserted-by":"publisher","first-page":"194","DOI":"10.1016\/j.jbiotec.2017.04.020","article-title":"The BRENDA enzyme information\nsystem-From a database to an expert system","volume":"261","author":"Schomburg","year":"2017","journal-title":"J. Biotechnol."},{"key":"2026082212130768300_cit46","doi-asserted-by":"publisher","first-page":"2187","DOI":"10.1007\/s10994-021-06023-5","article-title":"Density-based\nweighting\nfor imbalanced regression","volume":"110","author":"Steininger","year":"2021","journal-title":"Mach. Learn."},{"key":"2026082212130768300_cit47","article-title":"Reinforcement\nlearning: An introduction","volume-title":"A Bradford Book","author":"Sutton","year":"2018"},{"key":"2026082212130768300_cit48","doi-asserted-by":"publisher","first-page":"D638","DOI":"10.1093\/nar\/gkac1000","article-title":"The STRING database\nin 2023: protein-protein association networks and functional enrichment\nanalyses for any sequenced genome of interest","volume":"51","author":"Szklarczyk","year":"2023","journal-title":"Nucleic Acids Res."},{"key":"2026082212130768300_cit49","doi-asserted-by":"publisher","first-page":"737","DOI":"10.1101\/gr.214270.116","article-title":"Fast and accurate de\nnovo genome assembly from long uncorrected reads","volume":"27","author":"Vaser","year":"2017","journal-title":"Genome Res."},{"key":"2026082212130768300_cit50","doi-asserted-by":"publisher","first-page":"201","DOI":"10.1002\/yea.3278","article-title":"CRISPR\/Cpf1 enables\nfast and simple genome editing of Saccharomyces cerevisiae","volume":"35","author":"Verwaal","year":"2018","journal-title":"Yeast"},{"key":"2026082212130768300_cit51","unstructured":"Wick, R.; Volkening, J.\n          Porechop: adapter\ntrimmer for Oxford Nanopore reads. 2018, Available online: github com\/rrwick\/Porechop."},{"key":"2026082212130768300_cit52","unstructured":"Wright, C.; Wykes, M.\n          Medaka: sequence correction\nprovided by ONT research. 2023."},{"key":"2026082212130768300_cit53","doi-asserted-by":"publisher","first-page":"33","DOI":"10.1016\/j.ymben.2018.05.002","article-title":"Iterative\nalgorithm-guided\ndesign of massive strain libraries, applied to itaconic acid production\nin yeast","volume":"48","author":"Young","year":"2018","journal-title":"Metab. Eng."},{"key":"2026082212130768300_cit54","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-020-17910-1","article-title":"Combining mechanistic\nand machine learning models for predictive engineering and optimization\nof tryptophan metabolism","volume":"11","author":"Zhang","year":"2020","journal-title":"Nat. Commun."}],"container-title":["ACS Synthetic Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.acs.org\/doi\/pdf\/10.1021\/acssynbio.5c00864","content-type":"application\/pdf","content-version":"vor","intended-application":"unspecified"},{"URL":"https:\/\/pubs.acs.org\/asbcd6\/article-pdf\/15\/8\/3146\/66162443\/acssynbio.5c00864.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.acs.org\/asbcd6\/article-pdf\/15\/8\/3146\/66162443\/acssynbio.5c00864.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,22]],"date-time":"2026-08-22T16:13:13Z","timestamp":1787415193000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.acs.org\/asbcd6\/article\/15\/8\/3146\/5228142\/Machine-Learning-Assisted-Pathway-Optimization-in"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,22]]},"references-count":54,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2026,7,22]]},"published-print":{"date-parts":[[2026,8,21]]}},"URL":"https:\/\/doi.org\/10.1021\/acssynbio.5c00864","relation":{},"ISSN":["2161-5063"],"issn-type":[{"value":"2161-5063","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,22]]}}}