{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T09:50:02Z","timestamp":1791539402710,"version":"4.3.4"},"reference-count":58,"publisher":"Springer Science and Business Media LLC","license":[{"start":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T00:00:00Z","timestamp":1791504000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T00:00:00Z","timestamp":1791504000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Nat Comput Sci"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    The success of generative molecular design depends on the synthetic accessibility of the molecules it creates. While computational models can help predict chemical reactions, their current effectiveness is limited by a lack of reference data and the difficulty of accurately predicting reaction regioselectivity. Here, to address this, we present an integrated closed-loop workflow that combines efficient reaction data acquisition with high-accuracy deep learning. We developed an active learning strategy, driven by a tree-based ensemble, to guide the acquisition of the most informative laboratory experiments. This experimental campaign yielded a diverse benchmark for C\u2013H borylation. The bespoke dataset allowed us to train and evaluate geometric graph neural networks. We demonstrate that augmenting these symmetry-aware models with self-supervised auxiliary tasks consistently improves performance for both reaction outcome prediction and regioselectivity prediction. Prospective tests on unseen substrates, featuring challenging\n                    <jats:italic>N<\/jats:italic>\n                    -heteroaryl motifs, pinpointed the correct borylation positions in all cases, illustrating the practical utility of our workflow for efficient drug optimization.\n                  <\/jats:p>","DOI":"10.1038\/s43588-026-01063-0","type":"journal-article","created":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T09:02:22Z","timestamp":1791536542000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Advancing chemical reaction prediction in data-scarce drug discovery with active and geometric deep learning"],"prefix":"10.1038","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-0225-9107","authenticated-orcid":false,"given":"Mason","family":"Minot","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-7444-3946","authenticated-orcid":false,"given":"Yannick","family":"Stenzhorn","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7338-4103","authenticated-orcid":false,"given":"Jens","family":"Wolfard","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sebastian","family":"Strobel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Philippe","family":"Jablonski","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Daniel","family":"Zimmerli","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Martin","family":"Binder","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3164-9270","authenticated-orcid":false,"given":"Uwe","family":"Grether","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7895-497X","authenticated-orcid":false,"given":"Rainer E.","family":"Martin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8063-9952","authenticated-orcid":false,"given":"Alex T.","family":"M\u00fcller","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0346-3786","authenticated-orcid":false,"given":"David F.","family":"Nippa","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2628-1619","authenticated-orcid":false,"given":"Kenneth","family":"Atz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6706-1084","authenticated-orcid":false,"given":"Gisbert","family":"Schneider","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,10,9]]},"reference":[{"key":"1063_CR1","doi-asserted-by":"crossref","unstructured":"Nippa, D. F. et al. Expediting hit-to-lead progression in drug discovery through reaction prediction and multi-dimensional optimization. Nat. Commun. 16, 11646 (2025).","DOI":"10.1038\/s41467-025-66324-4"},{"key":"1063_CR2","doi-asserted-by":"publisher","first-page":"458","DOI":"10.1016\/j.drudis.2014.12.004","volume":"20","author":"D Reker","year":"2015","unstructured":"Reker, D. & Schneider, G. Active-learning strategies in computer-assisted drug discovery. Drug Discov. Today 20, 458\u2013465 (2015).","journal-title":"Drug Discov. Today"},{"key":"1063_CR3","doi-asserted-by":"publisher","first-page":"786","DOI":"10.1038\/s43588-024-00697-2","volume":"4","author":"D Van Tilborg","year":"2024","unstructured":"Van Tilborg, D. & Grisoni, F. Traversing chemical space with active deep learning for low-data drug discovery. Nat. Comput. Sci. 4, 786\u2013796 (2024).","journal-title":"Nat. Comput. Sci."},{"key":"1063_CR4","doi-asserted-by":"publisher","first-page":"105151","DOI":"10.1016\/j.engappai.2022.105151","volume":"115","author":"MA Ganaie","year":"2022","unstructured":"Ganaie, M. A., Hu, M., Malik, A. K., Tanveer, M. & Suganthan, P. N. Ensemble deep learning: a review. Eng. Appl. Artif. Intell. 115, 105151 (2022).","journal-title":"Eng. Appl. Artif. Intell."},{"key":"1063_CR5","doi-asserted-by":"crossref","unstructured":"Grinsztajn, L., Oyallon, E. & Varoquaux, G. Why do tree-based models still outperform deep learning on typical tabular data? In Proc. 36th International Conference on Neural Information Processing Systems (eds Koyejo, S. et al.) Vol. 35, 507\u2013520 (Curran Associates, 2022).","DOI":"10.52202\/068431-0037"},{"key":"1063_CR6","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1109\/MSP.2017.2693418","volume":"34","author":"MM Bronstein","year":"2017","unstructured":"Bronstein, M. M., Bruna, J., LeCun, Y., Szlam, A. & Vandergheynst, P. Geometric deep learning: going beyond Euclidean data. IEEE Signal Process. Mag. 34, 18\u201342 (2017).","journal-title":"IEEE Signal Process. Mag."},{"key":"1063_CR7","doi-asserted-by":"publisher","first-page":"256","DOI":"10.1038\/s42004-023-01047-5","volume":"6","author":"DF Nippa","year":"2023","unstructured":"Nippa, D. F. et al. Identifying opportunities for late-stage C\u2013H alkylation with high-throughput experimentation and in silico reaction screening. Commun. Chem. 6, 256 (2023).","journal-title":"Commun. Chem."},{"key":"1063_CR8","doi-asserted-by":"publisher","unstructured":"Jin, W. et al. Self-supervised learning on graphs: deep insights and new direction. Preprint at https:\/\/doi.org\/10.48550\/arXiv.2006.10141 (2020).","DOI":"10.48550\/arXiv.2006.10141"},{"key":"1063_CR9","unstructured":"Godwin, J. et al. Simple GNN regularisation for 3D molecular property prediction and beyond. In Proc. International Conference on Learning Representations https:\/\/openreview.net\/forum?id=1wVvweK3oIb (OpenReview.net, 2022)."},{"key":"1063_CR10","unstructured":"Zaidi, S. et al. Pre-training via denoising for molecular property prediction. In Proc. 11th International Conference on Learning Representations https:\/\/openreview.net\/forum?id=tYIMtogyee (OpenReview.net, 2023)."},{"key":"1063_CR11","doi-asserted-by":"publisher","first-page":"258","DOI":"10.2533\/chimia.2022.258","volume":"76","author":"DF Nippa","year":"2022","unstructured":"Nippa, D. F. et al. Late-stage functionalization and its impact on modern drug discovery: medicinal chemistry and chemical biology highlights. Chimia 76, 258\u2013258 (2022).","journal-title":"Chimia"},{"key":"1063_CR12","doi-asserted-by":"publisher","first-page":"864","DOI":"10.1021\/ar200206a","volume":"45","author":"JF Hartwig","year":"2012","unstructured":"Hartwig, J. F. Borylation and silylation of C\u2013H bonds: a platform for diverse C\u2013H bond functionalizations. Acc. Chem. Res. 45, 864\u2013873 (2012).","journal-title":"Acc. Chem. Res."},{"key":"1063_CR13","doi-asserted-by":"publisher","first-page":"22076","DOI":"10.1021\/jacs.3c04986","volume":"145","author":"E Caldeweyher","year":"2023","unstructured":"Caldeweyher, E. et al. Hybrid machine learning approach to predict the site selectivity of iridium-catalyzed arene borylation. J. Am. Chem. Soc. 145, 22076\u201322086 (2023).","journal-title":"J. Am. Chem. Soc."},{"key":"1063_CR14","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1023\/A:1010933404324","volume":"45","author":"L Breiman","year":"2001","unstructured":"Breiman, L. Random forests. Mach. Learn. 45, 5\u201332 (2001).","journal-title":"Mach. Learn."},{"key":"1063_CR15","unstructured":"Prokhorenkova, L., Gusev, G., Vorobev, A., Dorogush, A. V., & Gulin, A. CatBoost: unbiased boosting with categorical features. In Proc. Advances in Neural Information Processing Systems 31 (eds Bengio, S. et al.) (Curran Associates, 2018)."},{"key":"1063_CR16","doi-asserted-by":"crossref","unstructured":"Chen, T. & Guestrin, C. Xgboost: a scalable tree boosting system. In Proc. 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 785\u2013794 (ACM, 2016).","DOI":"10.1145\/2939672.2939785"},{"key":"1063_CR17","doi-asserted-by":"publisher","first-page":"742","DOI":"10.1021\/ci100050t","volume":"50","author":"D Rogers","year":"2010","unstructured":"Rogers, D. & Hahn, M. Extended-connectivity fingerprints. J. Chem. Inf. Model. 50, 742\u2013754 (2010).","journal-title":"J. Chem. Inf. Model."},{"key":"1063_CR18","doi-asserted-by":"publisher","unstructured":"Houlsby, N., Huszar, F., Ghahramani, Z. & Lengyel, M. Bayesian active learning for classification and preference learning. Preprint at https:\/\/doi.org\/10.48550\/arXiv.1112.5745 (2011).","DOI":"10.48550\/arXiv.1112.5745"},{"key":"1063_CR19","unstructured":"Kirsch, A. et al. Stochastic batch acquisition: a simple baseline for deep active learning. In Proc. Transactions on Machine Learning Research 2835\u20138856 https:\/\/openreview.net\/forum?id=vcHwQyNBjW (OpenReview.net, 2023)."},{"key":"1063_CR20","unstructured":"Ash, J. T., Zhang, C., Krishnamurthy, A., Langford, J. & Agarwal, A. Deep batch active learning by diverse, uncertain gradient lower bounds. In Proc. International Conference on Learning Representations https:\/\/openreview.net\/forum?id=ryghZJBKPS (OpenReview.net, 2020)."},{"key":"1063_CR21","doi-asserted-by":"publisher","unstructured":"Brochu, E., Cora, V. M. & De Freitas, N. A tutorial on Bayesian optimization of expensive cost functions, with application to active user modeling and hierarchical reinforcement learning. Peprint at https:\/\/doi.org\/10.48550\/arXiv.1012.2599 (2010).","DOI":"10.48550\/arXiv.1012.2599"},{"key":"1063_CR22","unstructured":"Srinivas, N., Krause, A., Kakade, S. M. & Seeger, M. Gaussian process optimization in the bandit setting: no regret and experimental design. Adv. Neural Inf. Process Syst. 21, 1689\u20131697 (2009)."},{"key":"1063_CR23","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s12864-019-6413-7","volume":"21","author":"D Chicco","year":"2020","unstructured":"Chicco, D. & Jurman, G. The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genomics 21, 1\u201313 (2020).","journal-title":"BMC Genomics"},{"key":"1063_CR24","unstructured":"Guo, C., Pleiss, G., Sun, Y. & Weinberger, K. Q. On calibration of modern neural networks. In Proc. 34th International Conference on Machine Learning (eds Precup, D. & Teh, Y. W.) Vol. 70 1321\u20131330 (PMLR, 2017)."},{"key":"1063_CR25","first-page":"2579","volume":"9","author":"L Van der Maaten","year":"2008","unstructured":"Van der Maaten, L. & Hinton, G. Visualizing data using t-SNE. J. Mach. Learn. Res. 9, 2579\u20132605 (2008).","journal-title":"J. Mach. Learn. Res."},{"key":"1063_CR26","unstructured":"M\u00fcller, A. T., Atz, K., Reutlinger, M. & Zorn, N. Combining graph attention and recurrent neural networks in a variational autoencoder for molecular representation learning and drug design. In Proc. ICML Workshop ML for Life and Material Science: From Theory to Industry Applications (OpenReview.net, 2024)."},{"key":"1063_CR27","first-page":"992","volume":"30","author":"K Sch\u00fctt","year":"2017","unstructured":"Sch\u00fctt, K. et al. SchNet: a continuous-filter convolutional neural network for modeling quantum interactions. Adv. Neural Inf. Process. Syst. 30, 992\u20131002 (2017).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"1063_CR28","unstructured":"Gasteiger, J., Giri, S., Margraf, J. T., & G\u00fcnnemann, S. Fast and uncertainty-aware directional message passing for non-equilibrium molecules. In Proc. NeurIPS Workshop on Machine Learning for Molecules (Curran Associates, 2020)."},{"key":"1063_CR29","unstructured":"Satorras, V. G., Hoogeboom, E. & Welling, M. E(n) equivariant graph neural networks. In Proc. 38th International Conference on Machine Learning (eds Meila, M. & Zhang, T.) Vol. 139 9323\u20139332 (PMLR, 2021)."},{"key":"1063_CR30","doi-asserted-by":"publisher","first-page":"239","DOI":"10.1038\/s41557-023-01360-5","volume":"16","author":"DF Nippa","year":"2024","unstructured":"Nippa, D. F. et al. Enabling late-stage drug diversification by high-throughput experimentation with geometric deep learning. Nat. Chem. 16, 239\u2013248 (2024).","journal-title":"Nat. Chem."},{"key":"1063_CR31","unstructured":"Sch\u00fctt, K., Unke, O. & Gasteigger, M. Equivariant message passing for the prediction of tensorial properties and molecular spectra. In Proc. 38th International Conference on Machine Learning (eds Meila, M. & Zhang, T.) 9377\u20139388 (PMLR, 2021)."},{"key":"1063_CR32","unstructured":"Bekkers, E. J., Vadgama, S., Hesselink, R. D., van der Linden, P. A. & Romero, D. W. Fast, expressive SE(n) equivariant networks through weight-sharing in position-orientation space. In Proc. International Conference on Learning Representations (OpenReview.net, 2024)."},{"key":"1063_CR33","doi-asserted-by":"publisher","unstructured":"Thomas, N. et al. Tensor field networks: rotation-and translation-equivariant neural networks for 3D point clouds. Preprint at https:\/\/doi.org\/10.48550\/arXiv.1802.08219 (2018).","DOI":"10.48550\/arXiv.1802.08219"},{"key":"1063_CR34","unstructured":"Brandstetter, J., Hesselink, R., van der Pol, E., Bekkers, E. J. & Welling, M. Geometric and physical quantities improve E(3) equivariant message passing. In Proc. International Conference on Learning Representations https:\/\/openreview.net\/forum?id=_xwr8gOBeV1 (OpenReview.net, 2022)."},{"key":"1063_CR35","unstructured":"Duval, A. A. et al. FAEnet: frame averaging equivariant GNN for materials modeling. In Proc. 40th International Conference on Machine Learning (eds Krause, A. et al.) 9013\u20139033 (PMLR, 2023)."},{"key":"1063_CR36","unstructured":"Passaro, S. & Zitnick, C. L. Reducing SO(3) convolutions to SO(2) for efficient equivariant GNNs. In Proc. 40th International Conference on Machine Learning (eds Krause, A. et al.) 27420\u201327438 (PMLR, 2023)."},{"key":"1063_CR37","doi-asserted-by":"publisher","first-page":"1376","DOI":"10.1021\/acs.jcim.1c01467","volume":"62","author":"J Lu","year":"2022","unstructured":"Lu, J. & Zhang, Y. Unified deep learning model for multitask reaction predictions with explanation. J. Chem. Inf. Model. 62, 1376\u20131387 (2022).","journal-title":"J. Chem. Inf. Model."},{"key":"1063_CR38","doi-asserted-by":"publisher","first-page":"3437","DOI":"10.1016\/S0040-4039(01)95429-2","volume":"20","author":"N Miyaura","year":"1979","unstructured":"Miyaura, N., Yamada, K. & Suzuki, A. A new stereospecific cross-coupling by the palladium-catalyzed reaction of 1-alkenylboranes with 1-alkenyl or 1-alkynyl halides. Tetrahedron Lett. 20, 3437\u20133440 (1979).","journal-title":"Tetrahedron Lett."},{"key":"1063_CR39","unstructured":"Minot, M. & Schneider, G. AIM: adaptive intervention for deep multi-task learning of molecular properties. In Proc. NeurIPS AI for Science Workshop https:\/\/openreview.net\/forum?id=LLTx78SGeM (OpenReview.net, 2025)."},{"key":"1063_CR40","doi-asserted-by":"publisher","first-page":"4","DOI":"10.1016\/j.cels.2023.12.003","volume":"15","author":"M Minot","year":"2024","unstructured":"Minot, M. & Reddy, S. T. Meta learning addresses noisy and under-labeled data in machine learning-guided antibody engineering. Cell Syst. 15, 4\u201318 (2024).","journal-title":"Cell Syst."},{"key":"1063_CR41","unstructured":"Malinin, A., Prokhorenkova, L. & Ustimenko, A. Uncertainty in gradient boosting via ensembles. In Proc. International Conference on Learning Representations https:\/\/openreview.net\/forum?id=1Jv6b0Zq3qi (OpenReview.net, 2021)."},{"key":"1063_CR42","unstructured":"Depeweg, S., Hernandez-Lobato, J-.M., Doshi-Velez, F. & Udluft, S. Decomposition of uncertainty in Bayesian deep learning for efficient and risk-sensitive learning. In Proc. of the 35th International Conference on Machine Learning (eds Dy, J. & Krause, A.) Vol. 80, 1184\u20131193 https:\/\/proceedings.mlr.press\/v80\/depeweg18a.html (PMLR, 2018)."},{"key":"1063_CR43","doi-asserted-by":"publisher","unstructured":"RDKit Development Team. RDKit: open-source cheminformatics software. Zenodo https:\/\/doi.org\/10.5281\/zenodo.7671152 (2022).","DOI":"10.5281\/zenodo.7671152"},{"key":"1063_CR44","unstructured":"Bergstra, J., Yamins, D. & Cox, D. Making a science of model search: hyperparameter optimization in hundreds of dimensions for vision architectures. In Proc. 30th International Conference on Machine Learning (eds Dasgupta, S. & McAllester, D.) Vol. 28, 115\u2013123 (PMLR, 2013)."},{"key":"1063_CR45","doi-asserted-by":"publisher","first-page":"e202400361","DOI":"10.1002\/minf.202400361","volume":"44","author":"DF Nippa","year":"2025","unstructured":"Nippa, D. F. et al. Simple user-friendly reaction format. Mol. Inform. 44, e202400361 (2025).","journal-title":"Mol. Inform."},{"key":"1063_CR46","doi-asserted-by":"publisher","first-page":"2887","DOI":"10.1021\/jm9602928","volume":"39","author":"GW Bemis","year":"1996","unstructured":"Bemis, G. W. & Murcko, M. A. The properties of known drugs. 1. Molecular frameworks. J. Med. Chem. 39, 2887\u20132893 (1996).","journal-title":"J. Med. Chem."},{"key":"1063_CR47","unstructured":"Liao, Y-.L., Wood, B., Das, A. & Smidt, T. EquiformerV2: improved equivariant transformer for scaling to higher-degree representations. In Proc. International Conference on Learning Representations https:\/\/openreview.net\/forum?id=mCOBKZmrzD (OpenReview.net, 2024)."},{"key":"1063_CR48","unstructured":"Paszke, A. et al. PyTorch: an imperative style, high-performance deep learning library. In Proc. Advances in Neural Information Processing Systems (eds Wallach, H. M. et al.) Vol. 32 8026\u20138037 (Curran Associates, 2019)."},{"key":"1063_CR49","unstructured":"Kingma, D. P., & Ba, J. Adam: a method for stochastic optimization. In Proc. 3rd International Conference on Learning Representations (eds Bengio, Y. & LeCun, Y.) (ICLR 2015)."},{"key":"1063_CR50","doi-asserted-by":"crossref","unstructured":"Hou, Z. et al. Graphmae: self-supervised masked graph autoencoders. In Proc. 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (eds Zhang, A. & Rangwala, H.) 594\u2013604 (ACM, 2022).","DOI":"10.1145\/3534678.3539321"},{"key":"1063_CR51","doi-asserted-by":"crossref","unstructured":"Devlin, J., Chang, M-.W., Lee, K. & Toutanova, K. BERT: pre-training of deep bidirectional transformers for language understanding. In Proc. Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (eds Burstein, J. et al.) Vol. 1 4171\u20134186 (ACL, 2019).","DOI":"10.18653\/v1\/N19-1423"},{"key":"1063_CR52","doi-asserted-by":"publisher","first-page":"10024","DOI":"10.1021\/ja00051a040","volume":"114","author":"AK Rapp\u00e9","year":"1992","unstructured":"Rapp\u00e9, A. K., Casewit, C. J., Colwell, K., Goddard III, W. A. & Skiff, W. M. UFF, a full periodic table force field for molecular mechanics and molecular dynamics simulations. J. Am. Chem. Soc. 114, 10024\u201310035 (1992).","journal-title":"J. Am. Chem. Soc."},{"key":"1063_CR53","doi-asserted-by":"crossref","unstructured":"Rasley, J., Rajbhandari, S., Ruwase, O. & He, Y. DeepSpeed: system optimizations enable training deep learning models with over 100 billion parameters. In Proc. 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 3505\u20133506 (ACM, 2020).","DOI":"10.1145\/3394486.3406703"},{"key":"1063_CR54","doi-asserted-by":"publisher","first-page":"747","DOI":"10.1021\/ci9803381","volume":"39","author":"D Butina","year":"1999","unstructured":"Butina, D. Unsupervised data base clustering based on daylight\u2019s fingerprint and Tanimoto similarity: a fast and automated way to cluster small and large data sets. J. Chem. Inf. Comput. Sci. 39, 747\u2013750 (1999).","journal-title":"J. Chem. Inf. Comput. Sci."},{"key":"1063_CR55","first-page":"2825","volume":"12","author":"F Pedregosa","year":"2011","unstructured":"Pedregosa, F. et al. Scikit-learn: machine learning in Python. J. Mach. Learn. Res. 12, 2825\u20132830 (2011).","journal-title":"J. Mach. Learn. Res."},{"key":"1063_CR56","doi-asserted-by":"publisher","first-page":"4354","DOI":"10.21105\/joss.04101","volume":"7","author":"NS Detlefsen","year":"2022","unstructured":"Detlefsen, N. S. et al. TorchMetrics\u2014a review-centric contributor-friendly library for PyTorch. J. Open Source Softw. 7, 4354 https:\/\/doi.org\/10.21105\/joss.04101 (2022).","journal-title":"J. Open Source Softw."},{"key":"1063_CR57","doi-asserted-by":"publisher","unstructured":"Minot, M. et al. Datasets for \u201cAdvancing chemical reaction prediction in data-scarce drug discovery with active and geometric deep learning\u201d. Zenodo https:\/\/doi.org\/10.5281\/zenodo.20773622 (2026).","DOI":"10.5281\/zenodo.20773622"},{"key":"1063_CR58","doi-asserted-by":"publisher","unstructured":"Minot, M. et al. Code and model weights for \u201cAdvancing chemical reaction prediction in data-scarce drug discovery with active and geometric deep learning\u201d. Zenodo https:\/\/doi.org\/10.5281\/zenodo.20783136 (2026).","DOI":"10.5281\/zenodo.20783136"}],"container-title":["Nature Computational Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s43588-026-01063-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s43588-026-01063-0","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s43588-026-01063-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T09:02:40Z","timestamp":1791536560000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s43588-026-01063-0"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,9]]},"references-count":58,"alternative-id":["1063"],"URL":"https:\/\/doi.org\/10.1038\/s43588-026-01063-0","relation":{},"ISSN":["2662-8457"],"issn-type":[{"value":"2662-8457","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,9]]},"assertion":[{"value":"23 September 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"1 September 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"9 October 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"G.S. declares a potential financial conflict of interest as a co-founder of inSili.com, Zurich and Xanadys, Zurich and in his role as a scientific consultant to the pharmaceutical industry. Y.S., J.W., S.S., P.J., D.Z., M.B., U.G., R.E.M., A.T.M., D.F.N. and K.A. are full employees of F. Hoffmann-La Roche. The other authors declare no competing interests.","order":1,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Ethics"}}]}}