{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,28]],"date-time":"2026-09-28T14:57:14Z","timestamp":1790607434697,"version":"4.1.0"},"reference-count":38,"publisher":"American Chemical Society (ACS)","issue":"18","license":[{"start":{"date-parts":[[2026,9,8]],"date-time":"2026-09-08T00:00:00Z","timestamp":1788825600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,9,8]],"date-time":"2026-09-08T00:00:00Z","timestamp":1788825600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,9,8]],"date-time":"2026-09-08T00:00:00Z","timestamp":1788825600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-045"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["21807088"],"award-info":[{"award-number":["21807088"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["22377098"],"award-info":[{"award-number":["22377098"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100012166","name":"National Key Research and Development Program of China","doi-asserted-by":"publisher","award":["2023YFA0914300"],"award-info":[{"award-number":["2023YFA0914300"]}],"id":[{"id":"10.13039\/501100012166","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/na","name":"Shaanxi Provincial Outstanding Young Scholars Fund","doi-asserted-by":"publisher","award":["2025JC-JCQN-048"],"award-info":[{"award-number":["2025JC-JCQN-048"]}],"id":[{"id":"10.13039\/na","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,9,28]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Identifying enzymes capable of catalyzing specific chemical transformations across large sequence databases remains a major challenge in biocatalyst discovery. Conventional fingerprint-based methods capture global molecular structure but fail to represent bond-breaking and bond-forming events, limiting generalization to structurally novel reactions. We introduce a dual-track evaluation framework to distinguish true generalization from memorization, assessing retrieval on structurally isolated reactions (n\u202f=\u202f50) within a 63,259-sequence enzyme pool. The strongest fingerprint baseline achieves R@10 = 0.020. To address this limitation, we develop GATv2-ECR, a heterogeneous dual-tower model integrating reaction-center graph encoding, a frozen ESM-2 sequence encoder, contrastive learning, and EC-aware soft reranking. GATv2-ECR achieves R@10 = 0.160 on isolated queries and R@10 = 0.308 on an out-of-distribution subset (n\u202f=\u202f39), capturing mechanistically relevant features and supporting generalizable enzyme retrieval under open-world conditions.<\/jats:p>","DOI":"10.1021\/acs.jcim.6c02414","type":"journal-article","created":{"date-parts":[[2026,9,8]],"date-time":"2026-09-08T07:01:57Z","timestamp":1788850917000},"page":"12196-12204","source":"Crossref","is-referenced-by-count":0,"title":["Mechanistically Informed Open-World Enzyme Retrieval\nwith a Dual-Tower Graph-Sequence Model"],"prefix":"10.1021","volume":"66","author":[{"given":"Siyuan","family":"Wang","sequence":"first","affiliation":[{"name":"Northwest University , , ,","place":["Xi\u2019an, China, 710127"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dan","family":"Wang","sequence":"additional","affiliation":[{"name":"Northwest University , , ,","place":["Xi\u2019an, China, 710127"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ying","family":"Ren","sequence":"additional","affiliation":[{"name":"Northwest A&F University , , Yangling, ,","place":["Xianyang, China, 712100"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8295-2854","authenticated-orcid":true,"given":"Xi","family":"Chen","sequence":"additional","affiliation":[{"name":"Northwest University , , ,","place":["Xi\u2019an, China, 710127"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"316","published-online":{"date-parts":[[2026,9,8]]},"reference":[{"issue":"1","key":"2026092810072676800_cit1","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1038\/s44160-021-00008-x","article-title":"Combining chemistry and protein engineering\nfor new-to-nature biocatalysis","volume":"1","author":"Miller","year":"2022","journal-title":"Nat. Synth."},{"issue":"1","key":"2026092810072676800_cit2","doi-asserted-by":"publisher","first-page":"12","DOI":"10.1021\/acscatal.5c04814","article-title":"Machine\nLearning-Driven Enzyme Mining","volume":"16","author":"Moorhoff","year":"2026","journal-title":"ACS Catal."},{"issue":"8057","key":"2026092810072676800_cit3","doi-asserted-by":"publisher","first-page":"146","DOI":"10.1038\/s41586-025-08592-0","article-title":"A compendium of\nhuman gene functions derived from evolutionary modelling","volume":"640","author":"Feuermann","year":"2025","journal-title":"Nature"},{"issue":"D1","key":"2026092810072676800_cit4","doi-asserted-by":"publisher","first-page":"D693","DOI":"10.1093\/nar\/gkab1016","article-title":"Rhea,\nthe reaction knowledgebase in 2022","volume":"50","author":"Bansal","year":"2022","journal-title":"Nucleic\nAcids Res."},{"key":"2026092810072676800_cit5","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1186\/s13321-026-01170-0","article-title":"Collision-free morgan fingerprints:\na principled approach to enhance machine learning performance and\ninterpretability in chemistry","volume":"18","author":"Li","year":"2026","journal-title":"J. Cheminf."},{"key":"2026092810072676800_cit6","doi-asserted-by":"publisher","first-page":"2434","DOI":"10.1021\/acs.jcim.5c02645","article-title":"ChemTorch: A Deep Learning Framework for Benchmarking and Developing\nChemical Reaction Property Prediction Models","volume":"66","author":"De Landsheere","year":"2025","journal-title":"J. Chem. Inf. Model."},{"issue":"4","key":"2026092810072676800_cit7","doi-asserted-by":"publisher","first-page":"2404671","DOI":"10.1002\/advs.202404671","article-title":"Precision adverse drug\nreactions prediction with heterogeneous graph\nneural network","volume":"12","author":"Gao","year":"2025","journal-title":"Adv. Sci."},{"key":"2026092810072676800_cit8","doi-asserted-by":"publisher","first-page":"115","DOI":"10.1016\/j.synbio.2025.11.012","article-title":"Transfer learning with pre-trained language models for protein expression\nlevel prediction in Escherichia coli","volume":"12","author":"Yang","year":"2026","journal-title":"Synth.\nSyst. Biotechnol."},{"issue":"1","key":"2026092810072676800_cit9","doi-asserted-by":"publisher","first-page":"2072","DOI":"10.1038\/s41467-025-57215-9","article-title":"CatPred: a comprehensive\nframework for deep learning\nin vitro enzyme kinetic parameters","volume":"16","author":"Boorla","year":"2025","journal-title":"Nat. Commun."},{"issue":"8090","key":"2026092810072676800_cit10","doi-asserted-by":"publisher","first-page":"639","DOI":"10.1038\/s41586-025-09697-2","article-title":"Enzyme specificity\nprediction using\ncross-attention graph neural networks","volume":"647","author":"Cui","year":"2025","journal-title":"Nature"},{"issue":"6639","key":"2026092810072676800_cit11","doi-asserted-by":"publisher","first-page":"1358","DOI":"10.1126\/science.adf2465","article-title":"Enzyme function\nprediction using contrastive learning","volume":"379","author":"Yu","year":"2023","journal-title":"Science"},{"issue":"Suppl S2","key":"2026092810072676800_cit12","doi-asserted-by":"publisher","first-page":"2263","DOI":"10.1007\/s10462-023-10577-2","article-title":"Attention-based\ngraph neural networks: a survey","volume":"56","author":"Sun","year":"2023","journal-title":"Artif. Intell.\nRe."},{"issue":"9","key":"2026092810072676800_cit13","doi-asserted-by":"publisher","first-page":"1445","DOI":"10.1101\/gr.279132.124","article-title":"Parameter-efficient\nfine-tuning on large protein language models improves signal peptide\nprediction","volume":"34","author":"Zeng","year":"2024","journal-title":"Genome Res."},{"key":"2026092810072676800_cit14","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2309.13944","article-title":"Provable training\nfor graph contrastive learning","author":"Yu","year":"2023","journal-title":"arXiv"},{"key":"2026092810072676800_cit15","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2009.13862","article-title":"Where\nis the model looking at?\u2013Concentrate and explain the network\nattention","author":"Xu","journal-title":"arXiv"},{"issue":"1","key":"2026092810072676800_cit16","doi-asserted-by":"publisher","first-page":"87","DOI":"10.1186\/s13321-022-00664-x","article-title":"Visualizing chemical space networks\nwith RDKit and NetworkX","volume":"14","author":"Scalfani","year":"2022","journal-title":"J. Cheminf."},{"issue":"34","key":"2026092810072676800_cit17","doi-asserted-by":"publisher","first-page":"8766","DOI":"10.1021\/acs.jpclett.4c02129","article-title":"Cross-modal\nprediction of spectral and structural descriptors via a pretrained\nmodel enhanced with chemical insights","volume":"15","author":"Yang","year":"2024","journal-title":"J. Phys.\nChem. Lett."},{"issue":"6","key":"2026092810072676800_cit18","doi-asserted-by":"publisher","first-page":"3089","DOI":"10.1021\/acs.chemrev.2c00798","article-title":"A brief\nintroduction to chemical reaction optimization","volume":"123","author":"Taylor","year":"2023","journal-title":"Chem. Rev."},{"issue":"6673","key":"2026092810072676800_cit19","doi-asserted-by":"publisher","DOI":"10.1126\/science.adh8615","article-title":"From nature\nto industry: Harnessing\nenzymes for biocatalysis","volume":"382","author":"Buller","year":"2023","journal-title":"Science"},{"issue":"5","key":"2026092810072676800_cit20","doi-asserted-by":"publisher","first-page":"bbaf491","DOI":"10.1093\/bib\/bbaf491","article-title":"A survey on deep learning\nfor drug-target\nbinding prediction: models, benchmarks, evaluation, and case studies","volume":"26","author":"Debnath","year":"2025","journal-title":"Briefings Bioinf."},{"issue":"1","key":"2026092810072676800_cit21","doi-asserted-by":"publisher","first-page":"8211","DOI":"10.1038\/s41467-023-44113-1","article-title":"UniKP: a unified\nframework for the prediction of enzyme kinetic parameters","volume":"14","author":"Yu","year":"2023","journal-title":"Nat. Commun."},{"key":"2026092810072676800_cit22","doi-asserted-by":"publisher","first-page":"7741","DOI":"10.1109\/JBHI.2025.3567518","article-title":"TSLAmy: a\nnovel amyloid hexapeptide aggregation prediction\napproach based on two-stage learning","volume":"29","author":"Huang","year":"2025","journal-title":"IEEE J.\nBiomed. Health Inform."},{"issue":"12","key":"2026092810072676800_cit23","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2520070123","article-title":"Dual-encoder contrastive\nlearning accelerates enzyme discovery","volume":"123","author":"Rocks","year":"2026"},{"key":"2026092810072676800_cit24","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2408.13659","article-title":"Reactzyme: A benchmark for enzyme-reaction\nprediction","author":"Hua","year":"2024","journal-title":"arXiv"},{"issue":"1","key":"2026092810072676800_cit25","doi-asserted-by":"publisher","first-page":"39","DOI":"10.1186\/s13321-020-00443-6","article-title":"QSAR-derived\naffinity fingerprints (part 1): fingerprint construction and modeling\nperformance for similarity searching, bioactivity classification and\nscaffold hopping","volume":"12","author":"S\u030ckuta","year":"2020","journal-title":"J. Cheminf."},{"issue":"6","key":"2026092810072676800_cit26","doi-asserted-by":"publisher","first-page":"1169","DOI":"10.1021\/acs.jcim.7b00656","article-title":"Multiclassification prediction of enzymatic reactions for oxidoreductases\nand hydrolases using reaction fingerprints and machine learning methods","volume":"58","author":"Cai","year":"2018","journal-title":"J. Chem. Inf. Model."},{"issue":"1","key":"2026092810072676800_cit27","doi-asserted-by":"publisher","first-page":"1654","DOI":"10.1038\/s41467-025-56827-5","article-title":"Learning and actioning general principles\nof cancer cell drug sensitivity","volume":"16","author":"Carli","year":"2025","journal-title":"Nat. Commun."},{"key":"2026092810072676800_cit28","doi-asserted-by":"publisher","first-page":"208","DOI":"10.1007\/s44443-026-00611-y","article-title":"PRISM-Rec:\nprogressive relational interest and semantic\nmodeling for knowledge graph-enhanced recommendation","volume":"38","author":"Pasupuleti","year":"2026","journal-title":"J. King Saud Univ. Comput. Inf. Sci."},{"issue":"23","key":"2026092810072676800_cit29","doi-asserted-by":"publisher","first-page":"2309075","DOI":"10.1002\/smll.202309075","article-title":"Dual-track multifunctional bimetallic\nmetal-organic frameworks for antibiotic enrichment and detection","volume":"20","author":"Dong","year":"2024","journal-title":"Small"},{"issue":"1","key":"2026092810072676800_cit30","doi-asserted-by":"publisher","first-page":"1690","DOI":"10.1038\/s42003-024-07359-z","article-title":"Improved enzyme functional\nannotation prediction using contrastive learning with structural inference","volume":"7","author":"Yang","year":"2024","journal-title":"Commun. Biol."},{"key":"2026092810072676800_cit31","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41746-025-02240-7","article-title":"The structure-preserving\nspectral graph neural network for dual kinase\ninhibitors and synergy scoring in gastric cancer","volume":"9","author":"Zhang","year":"2025","journal-title":"npj Digit. Med."},{"key":"2026092810072676800_cit32","doi-asserted-by":"publisher","first-page":"613","DOI":"10.1007\/s11831-025-10317-5","article-title":"Advancements\nin Molecular Property\nPrediction: A Survey of Single and Multimodal Approaches","volume":"33","author":"Liyaqat","year":"2025","journal-title":"Arch. Computat. Methods Eng."},{"issue":"6","key":"2026092810072676800_cit33","doi-asserted-by":"publisher","first-page":"bbad358","DOI":"10.1093\/bib\/bbad358","article-title":"Generative\nmodels for\nprotein sequence modeling: recent advances and future directions","volume":"24","author":"Mardikoraem","year":"2023","journal-title":"Briefings Bioinf."},{"issue":"11","key":"2026092810072676800_cit34","doi-asserted-by":"publisher","first-page":"15051","DOI":"10.1109\/TNNLS.2023.3283523","article-title":"Challenges and opportunities in deep\nreinforcement learning with graph neural networks: A comprehensive\nreview of algorithms and applications","volume":"35","author":"Munikoti","year":"2024","journal-title":"IEEE\nTrans. Neural Netw. Learn. Syst"},{"issue":"17","key":"2026092810072676800_cit35","doi-asserted-by":"publisher","first-page":"4674","DOI":"10.1016\/j.cell.2025.06.014","article-title":"Advancing\nprotein evolution with inverse folding models integrating structural\nand evolutionary constraints","volume":"188","author":"Fei","year":"2025","journal-title":"Cell"},{"issue":"55","key":"2026092810072676800_cit36","doi-asserted-by":"publisher","DOI":"10.1002\/chem.202401626","article-title":"Computer-aided\nsynthesis planning\n(CASP) and machine learning: optimizing chemical reaction conditions","volume":"30","author":"Han","year":"2024","journal-title":"Chem. -Eur. J."},{"key":"2026092810072676800_cit37","doi-asserted-by":"publisher","first-page":"2133","DOI":"10.1038\/s41467-026-68776-8","article-title":"Stereoselective cyano\ntranslocation reaction enabled by photoenzymatic catalysis.","volume":"17","author":"Duan","year":"2026","journal-title":"Nat. Commun."},{"issue":"3","key":"2026092810072676800_cit38","doi-asserted-by":"publisher","first-page":"579","DOI":"10.1080\/01614940.2024.2320802","article-title":"Michael reaction in organocascade\ncatalysis: A powerful\ntool for creating architectural complexity in the construction of\nmedicinally privileged heterocyclic scaffolds","volume":"67","author":"Yadav","year":"2025","journal-title":"Catal. Rev."}],"container-title":["Journal of Chemical Information and Modeling"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.acs.org\/doi\/pdf\/10.1021\/acs.jcim.6c02414","content-type":"application\/pdf","content-version":"vor","intended-application":"unspecified"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/66\/18\/12196\/68437789\/acs.jcim.6c02414.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article-pdf\/66\/18\/12196\/68437789\/acs.jcim.6c02414.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,28]],"date-time":"2026-09-28T14:12:06Z","timestamp":1790604726000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.acs.org\/jcisd8\/article\/66\/18\/12196\/5425803\/Mechanistically-Informed-Open-World-Enzyme"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,8]]},"references-count":38,"journal-issue":{"issue":"18","published-online":{"date-parts":[[2026,9,8]]},"published-print":{"date-parts":[[2026,9,28]]}},"URL":"https:\/\/doi.org\/10.1021\/acs.jcim.6c02414","relation":{},"ISSN":["1549-9596","1549-960X"],"issn-type":[{"value":"1549-9596","type":"print"},{"value":"1549-960X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,8]]}}}