{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T11:45:01Z","timestamp":1785498301346,"version":"3.56.0"},"reference-count":41,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,3,14]],"date-time":"2026-03-14T00:00:00Z","timestamp":1773446400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,4,23]],"date-time":"2026-04-23T00:00:00Z","timestamp":1776902400000},"content-version":"vor","delay-in-days":40,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J Cheminform"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Selecting an effective docking algorithm is highly context-dependent, and no single method performs reliably across structural, chemical, and protocol regimes. MolAS is a lightweight algorithm-selection model that predicts per-algorithm performance from pretrained protein and ligand embeddings using attentional pooling and a shallow residual decoder. With hundreds to a few thousand labelled complexes, MolAS achieves up to a 15 percentage-point absolute improvement over the single\u00a0best solver (SBS) and closes 17\u201366% of the virtual best solver (VBS)\u2013SBS gap across five docking benchmarks. Analyses of selection frequencies, margin-conditioned reliability, and benchmark-level oracle structure indicate that MolAS is most effective when the workflow-defined oracle landscape has low winner entropy and a reasonably separable top-solver region, but degrades under protocol mismatch that shifts solver rankings and changes the induced labels. These results suggest that, in the evaluated regime, robustness is limited less by representational capacity than by workflow- and protocol-induced instability in solver hierarchies, positioning MolAS as an in-domain selector for fixed pipelines and as a diagnostic tool for assessing when docking algorithm selection is well-posed.<\/jats:p>\n                  <jats:p>\n                    <jats:bold>Scientific Contribution<\/jats:bold>\n                    : MolAS introduces a controlled, embedding-based selector that reduces dependence on heavy graph encoders, enabling a cleaner separation between representational choices and workflow-defined label structure. A cross-benchmark and cross-protocol analysis links selection success and failure to oracle entropy, near-ties among top solvers, and protocol-induced ranking shifts, providing an evidence-backed diagnostic account of when docking algorithm selection is likely to yield gains. The findings differentiate this work from prior docking AS studies that report in-domain improvements under a single fixed workflow by explicitly characterising protocol dependence and motivating protocol-aware modelling as a route to stronger generalisation.\n                  <\/jats:p>\n                  <jats:p>\n                    <jats:bold>Graphical Abstract<\/jats:bold>\n                  <\/jats:p>","DOI":"10.1186\/s13321-026-01168-8","type":"journal-article","created":{"date-parts":[[2026,3,14]],"date-time":"2026-03-14T18:02:27Z","timestamp":1773511347000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Molecular embedding-based algorithm selection in protein-ligand docking"],"prefix":"10.1186","volume":"18","author":[{"given":"Jiabao Brad","family":"Wang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Siyuan","family":"Cao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hongxuan","family":"Wu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yiliang","family":"Yuan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mustafa","family":"M\u0131s\u0131r","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,3,14]]},"reference":[{"issue":"2","key":"1168_CR1","doi-asserted-by":"publisher","first-page":"83","DOI":"10.1007\/s40484-019-0172-y","volume":"7","author":"J Fan","year":"2019","unstructured":"Fan J, Ailing F, Zhang L (2019) Progress in molecular docking. Quant Biol 7(2):83\u201389. https:\/\/doi.org\/10.1007\/s40484-019-0172-y","journal-title":"Quant Biol"},{"issue":"1","key":"1168_CR2","doi-asserted-by":"publisher","first-page":"15451","DOI":"10.1038\/s41598-017-15571-7","volume":"7","author":"NM Hassan","year":"2017","unstructured":"Hassan NM, Alhossary AA, Mu Y, Kwoh C-K (2017) Protein-ligand blind docking using quickvina-w with inter-process spatio-temporal integration. Sci Rep 7(1):15451. https:\/\/doi.org\/10.1038\/s41598-017-15571-7","journal-title":"Sci Rep"},{"issue":"8","key":"1168_CR3","doi-asserted-by":"publisher","first-page":"1893","DOI":"10.1021\/ci300604z","volume":"53","author":"DR Koes","year":"2013","unstructured":"Koes DR, Baumgartner MP, Camacho CJ (2013) Lessons learned in empirical scoring with smina from the csar 2011 benchmarking exercise. J Chem Inf Model 53(8):1893\u20131904. https:\/\/doi.org\/10.1021\/ci300604z","journal-title":"J Chem Inf Model"},{"issue":"2","key":"1168_CR4","doi-asserted-by":"publisher","first-page":"455","DOI":"10.1002\/jcc.21334","volume":"31","author":"O Trott","year":"2010","unstructured":"Trott O, Olson AJ, vina A (2010) Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem 31(2):455\u2013461. https:\/\/doi.org\/10.1002\/jcc.21334","journal-title":"J Comput Chem"},{"issue":"9","key":"1168_CR5","doi-asserted-by":"publisher","first-page":"1169","DOI":"10.1093\/bioinformatics\/btq112","volume":"26","author":"PJ Ballester","year":"2010","unstructured":"Ballester PJ, Mitchell JBO (2010) A machine learning approach to predicting protein-ligand binding affinity with applications to molecular docking. Bioinformatics 26(9):1169\u20131175. https:\/\/doi.org\/10.1093\/bioinformatics\/btq112","journal-title":"Bioinformatics"},{"issue":"1","key":"1168_CR6","doi-asserted-by":"publisher","first-page":"193","DOI":"10.1186\/1471-2105-11-193","volume":"11","author":"T Cheng","year":"2010","unstructured":"Cheng T, Liu Z, Wang R (2010) A knowledge-guided strategy for improving the accuracy of scoring functions in binding affinity prediction. BMC Bioinform 11(1):193. https:\/\/doi.org\/10.1186\/1471-2105-11-193","journal-title":"BMC Bioinform"},{"issue":"2","key":"1168_CR7","doi-asserted-by":"publisher","first-page":"146","DOI":"10.2174\/157340911795677602","volume":"7","author":"X-Y Meng","year":"2011","unstructured":"Meng X-Y, Zhang H-X, Mezei M, Cui M (2011) Molecular docking: a powerful approach for structure-based drug discovery. Curr Comput Aided Drug Des 7(2):146\u2013157. https:\/\/doi.org\/10.2174\/157340911795677602","journal-title":"Curr Comput Aided Drug Des"},{"issue":"2","key":"1168_CR8","doi-asserted-by":"publisher","first-page":"895","DOI":"10.1021\/acs.jcim.8b00545","volume":"59","author":"M Su","year":"2018","unstructured":"Su M, Yang Q, Du Y, Feng G, Liu Z, Li Y, Wang R (2018) Comparative assessment of scoring functions: the casf-2016 update. J Chem Inf Model 59(2):895\u2013913. https:\/\/doi.org\/10.1021\/acs.jcim.8b00545","journal-title":"J Chem Inf Model"},{"key":"1168_CR9","unstructured":"Corso G, St\u00e4rk H, Jing B (2022) Regina Barzilay, and Tommi Jaakkola. Diffdock: Diffusion steps, twists, and turns for molecular docking. arXiv:2210.01776"},{"issue":"2","key":"1168_CR10","doi-asserted-by":"publisher","first-page":"310","DOI":"10.1038\/s41592-024-02516-y","volume":"22","author":"D Cao","year":"2025","unstructured":"Cao D, Chen M, Zhang R, Wang Z, Huang M, Jie Yu, Jiang X, Fan Z, Zhang W, Zhou H et al (2025) Surfdock is a surface-informed diffusion generative model for reliable and accurate protein-ligand complex prediction. Nat Methods 22(2):310\u2013322. https:\/\/doi.org\/10.1038\/s41592-024-02516-y","journal-title":"Nat Methods"},{"issue":"1","key":"1168_CR11","doi-asserted-by":"publisher","first-page":"67","DOI":"10.1109\/4235.585893","volume":"1","author":"DH Wolpert","year":"2002","unstructured":"Wolpert DH, Macready WG (2002) No free lunch theorems for optimization. IEEE Trans Evol Comput 1(1):67\u201382. https:\/\/doi.org\/10.1109\/4235.585893","journal-title":"IEEE Trans Evol Comput"},{"key":"1168_CR12","doi-asserted-by":"publisher","unstructured":"Yuan Y, Misir M (2024a) Algorithm selection on molecular docking for state-of-the-art performance. In Proceedings of the 2024 8th International Conference on Computer Science and Artificial Intelligence, 295\u2013302. https:\/\/doi.org\/10.1145\/3709026.3709046","DOI":"10.1145\/3709026.3709046"},{"key":"1168_CR13","first-page":"65","volume-title":"Advances in computers","author":"JR Rice","year":"1976","unstructured":"Rice JR (1976) The algorithm selection problem. In: Rubinoff M, Yovits MC (eds) Advances in computers. Elsevier, Amsterdam, pp 65\u2013118"},{"issue":"1","key":"1168_CR14","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1162\/evco_a_00242","volume":"27","author":"P Kerschke","year":"2019","unstructured":"Kerschke P, Hoos HH, Neumann F, Trautmann H (2019) Automated algorithm selection: survey and perspectives. Evol Comput 27(1):3\u201345. https:\/\/doi.org\/10.1162\/evco_a_00242","journal-title":"Evol Comput"},{"key":"1168_CR15","doi-asserted-by":"publisher","unstructured":"Thornton C, Hutter F, Hoos HH, Leyton-Brown K (2013) Auto-weka: Combined selection and hyperparameter optimization of classification algorithms. In Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining. 847\u2013855. https:\/\/doi.org\/10.1145\/2487575.2487629","DOI":"10.1145\/2487575.2487629"},{"key":"1168_CR16","unstructured":"Feurer M, Klein A, Eggensperger K, Springenberg J, Blum M, Hutter F (2015) Efficient and robust automated machine learning. Advances in neural information processing systems. 28. https:\/\/proceedings.neurips.cc\/paper_files\/paper\/2015\/file\/11d0e6287202fced83f79975ec59a3a6-Paper.pdf"},{"issue":"1","key":"1168_CR17","doi-asserted-by":"publisher","first-page":"285","DOI":"10.1007\/s10994-017-5685-x","volume":"107","author":"I Olier","year":"2018","unstructured":"Olier I, Sadawi N, Bickerton GR, Vanschoren J, Grosan C, Soldatova L, King RD (2018) Meta-qsar: a large-scale application of meta-learning to drug design and discovery. Mach Learn 107(1):285\u2013311. https:\/\/doi.org\/10.1007\/s10994-017-5685-x","journal-title":"Mach Learn"},{"issue":"1","key":"1168_CR18","doi-asserted-by":"publisher","first-page":"8219","DOI":"10.1038\/s41598-023-35132-5","volume":"13","author":"T Chen","year":"2023","unstructured":"Chen T, Shu X, Zhou H, Beckford FA, Misir M (2023) Algorithm selection for protein-ligand docking: strategies and analysis on ace. Sci Rep 13(1):8219. https:\/\/doi.org\/10.1038\/s41598-023-35132-5","journal-title":"Sci Rep"},{"key":"1168_CR19","doi-asserted-by":"publisher","first-page":"291","DOI":"10.1016\/j.artint.2016.12.001","volume":"244","author":"M M\u0131s\u0131r","year":"2017","unstructured":"M\u0131s\u0131r M, Sebag M (2017) Alors: an algorithm recommender system. Artif Intell 244:291\u2013314. https:\/\/doi.org\/10.1016\/j.artint.2016.12.001","journal-title":"Artif Intell"},{"issue":"1","key":"1168_CR20","doi-asserted-by":"publisher","first-page":"164","DOI":"10.1007\/s44371-025-00246-4","volume":"2","author":"AOH Zayed","year":"2025","unstructured":"Zayed AOH (2025) Optimizing protein-ligand docking through machine learning: algorithm selection with autodock vina. Discov Chem 2(1):164. https:\/\/doi.org\/10.1007\/s44371-025-00246-4","journal-title":"Discov Chem"},{"key":"1168_CR21","doi-asserted-by":"crossref","unstructured":"Cao S, Wu H, Wang JB, Yuan Y, Misir M (2025b) Mc-gnnas-dock: Multi-criteria gnn-based algorithm selection for molecular docking. arXiv:2509.26377","DOI":"10.1007\/978-981-95-7075-1_46"},{"key":"1168_CR22","doi-asserted-by":"publisher","unstructured":"Yuan Y, Misir M (2024b) Enhancing molecular docking performance with a gnn-based algorithm selection model. In Proceedings of the 2024 8th International Conference on Computational Biology and Bioinformatics, 14\u201319. https:\/\/doi.org\/10.1145\/3715020.3715040","DOI":"10.1145\/3715020.3715040"},{"key":"1168_CR23","unstructured":"Yuan Y, Misir M (2024c) Gnnas-dock: Budget aware algorithm selection with graph neural networks for molecular docking. In NeurIPS 2024 Workshop on AI for New Drug Modalities . https:\/\/openreview.net\/forum?id=W4d1a4odSc"},{"key":"1168_CR24","unstructured":"ESM Team (2024) Esm cambrian: Revealing the mysteries of proteins with unsupervised learning. https:\/\/evolutionaryscale.ai\/blog\/esm-cambrian"},{"key":"1168_CR25","unstructured":"Ahmad W, Simon E, Chithrananda S, Grand G, Ramsundar B (2022) Chemberta-2: Towards chemical foundation models. arXiv:2209.01712"},{"key":"1168_CR26","unstructured":"Corso G, Deng A, Fry B, Polizzi N, Barzilay R, Jaakkola T (2024) Deep confident steps to new pockets: Strategies for docking generalization. arXiv:2402.18396"},{"issue":"1","key":"1168_CR27","doi-asserted-by":"publisher","first-page":"43","DOI":"10.1186\/s13321-021-00522-2","volume":"13","author":"AT McNutt","year":"2021","unstructured":"McNutt AT, Francoeur P, Aggarwal R, Masuda T, Meli R, Ragoza M, Sunseri J, Koes DR (2021) Gnina 1.0: molecular docking with deep learning. J Cheminform 13(1):43. https:\/\/doi.org\/10.1186\/s13321-021-00522-2","journal-title":"J Cheminform"},{"key":"1168_CR28","doi-asserted-by":"publisher","unstructured":"Alcaide E, Gao Z, Ke G, Li Y, Zhang L, Zheng H, Zhou G (2025) Uni-mol docking v2: Towards realistic and accurate binding pose prediction. In International Conference on Artificial Neural Networks. Springer. 34\u201341. https:\/\/doi.org\/10.1007\/978-3-032-04552-2_5","DOI":"10.1007\/978-3-032-04552-2_5"},{"issue":"9","key":"1168_CR29","doi-asserted-by":"publisher","first-page":"789","DOI":"10.1038\/s43588-023-00511-5","volume":"3","author":"X Zhang","year":"2023","unstructured":"Zhang X, Zhang O, Shen C, Wanglin Q, Chen S, Hanqun Cao Yu, Kang ZW, Wang E, Zhang J et al (2023) Efficient and accurate large library ligand docking with karmadock. Nat Comput Sci 3(9):789\u2013804. https:\/\/doi.org\/10.1038\/s43588-023-00511-5","journal-title":"Nat Comput Sci"},{"issue":"2","key":"1168_CR30","doi-asserted-by":"publisher","first-page":"302","DOI":"10.1021\/acs.accounts.6b00491","volume":"50","author":"Z Liu","year":"2017","unstructured":"Liu Z, Minyi S, Han L, Liu J, Yang Q, Li Y, Wang R (2017) Forging the basis for developing protein-ligand interaction scoring functions. Acc Chem Res 50(2):302\u2013309. https:\/\/doi.org\/10.1021\/acs.accounts.6b00491","journal-title":"Acc Chem Res"},{"issue":"3","key":"1168_CR31","doi-asserted-by":"publisher","first-page":"333","DOI":"10.1002\/prot.20512","volume":"60","author":"L Hu","year":"2005","unstructured":"Hu L, Benson ML, Smith RD, Lerner MG, Carlson HA (2005) Binding moad (mother of all databases). Proteins Struct Funct Bioinform 60(3):333\u2013340. https:\/\/doi.org\/10.1002\/prot.20512","journal-title":"Proteins Struct Funct Bioinform"},{"key":"1168_CR32","unstructured":"Jiang Y, Li X, Zhang Y, Han J, Xu Y, Pandit A, Zhang Z, Wang M, Wang M, Liu C, Yang G, Choi Y, Li W-J, Fu T, Wu F, Liu J (2025) Posex: AI defeats physics-based methods on protein ligand cross-docking. https:\/\/openreview.net\/forum?id=qHOU1aAyXA"},{"issue":"9","key":"1168_CR33","doi-asserted-by":"publisher","first-page":"3130","DOI":"10.1039\/D3SC04185A","volume":"15","author":"M Buttenschoen","year":"2024","unstructured":"Buttenschoen M, Morris GM, Deane CM (2024) Posebusters: Ai-based docking methods fail to generate physically valid poses or generalise to novel sequences. Chem Sci 15(9):3130\u20133139. https:\/\/doi.org\/10.1039\/D3SC04185A","journal-title":"Chem Sci"},{"issue":"4","key":"1168_CR34","doi-asserted-by":"publisher","first-page":"726","DOI":"10.1021\/jm061277y","volume":"50","author":"MJ Hartshorn","year":"2007","unstructured":"Hartshorn MJ, Verdonk ML, Chessari G, Brewerton SC, Mooij WTM, Mortenson PN, Murray CW (2007) Diverse, high-quality test set for the validation of protein- ligand docking performance. J Med Chem 50(4):726\u2013741. https:\/\/doi.org\/10.1021\/jm061277y","journal-title":"J Med Chem"},{"issue":"9","key":"1168_CR35","doi-asserted-by":"publisher","first-page":"4200","DOI":"10.1021\/acs.jcim.0c00411","volume":"60","author":"PG Francoeur","year":"2020","unstructured":"Francoeur PG, Masuda T, Sunseri J, Jia A, Iovanisci RB, Snyder I, Koes DR (2020) Three-dimensional convolutional neural networks and a cross-docked data set for structure-based drug design. J Chem Inf Model 60(9):4200\u20134215. https:\/\/doi.org\/10.1021\/acs.jcim.0c00411","journal-title":"J Chem Inf Model"},{"issue":"12","key":"1168_CR36","doi-asserted-by":"publisher","first-page":"5349","DOI":"10.1109\/TNNLS.2020.2966319","volume":"31","author":"F He","year":"2020","unstructured":"He F, Liu T, Tao D (2020) Why resnet works? Residuals generalize. IEEE Trans Neural Netw Learn Syst 31(12):5349\u20135362. https:\/\/doi.org\/10.1109\/TNNLS.2020.2966319","journal-title":"IEEE Trans Neural Netw Learn Syst"},{"key":"1168_CR37","doi-asserted-by":"publisher","unstructured":"Lin Z, Mishra C, Santiagom, Gong J, Thomas N, Mathur I, Tina-z jia, Sercu T, Chan S, Candido S, MacCarley J, Qureshi I, Safer J, Verkuil R, Rivera V, Blenderwang, Carolynkim, Halilakin (2025) evolutionaryscale\/esm: v3.2.3. https:\/\/doi.org\/10.5281\/zenodo.17353381","DOI":"10.5281\/zenodo.17353381"},{"issue":"7","key":"1168_CR38","doi-asserted-by":"publisher","first-page":"2496","DOI":"10.1021\/acs.jcim.3c01208","volume":"64","author":"R Gorantla","year":"2023","unstructured":"Gorantla R, Kubincov\u00e1 A, Wei\u00dfe AY, Mey ASJS (2023) From proteins to ligands: decoding deep learning methods for binding affinity prediction. J Chem Inf Model 64(7):2496\u20132507","journal-title":"J Chem Inf Model"},{"issue":"3","key":"1168_CR39","doi-asserted-by":"publisher","first-page":"947","DOI":"10.1021\/acs.jcim.8b00712","volume":"59","author":"J Sieg","year":"2019","unstructured":"Sieg J, Flachsenberg F, Rarey M (2019) In need of bias control: evaluating chemical data for machine learning in structure-based virtual screening. J Chem Inf Model 59(3):947\u2013961","journal-title":"J Chem Inf Model"},{"issue":"7","key":"1168_CR40","doi-asserted-by":"publisher","first-page":"2323","DOI":"10.1021\/acs.jcim.3c00771","volume":"64","author":"G Mqawass","year":"2024","unstructured":"Mqawass G, Popov P (2024) Graphlambda: fusion graph neural networks for binding affinity prediction. J Chem Inf Model 64(7):2323\u20132330. https:\/\/doi.org\/10.1021\/acs.jcim.3c00771","journal-title":"J Chem Inf Model"},{"key":"1168_CR41","doi-asserted-by":"publisher","first-page":"101894","DOI":"10.1016\/j.swevo.2025.101894","volume":"94","author":"G Cenikj","year":"2025","unstructured":"Cenikj G, Petelin G, Seiler M, Cenikj N, Eftimov T (2025) Landscape features in single-objective continuous optimization: have we hit a wall in algorithm selection generalization? Swarm Evol Comput 94:101894","journal-title":"Swarm Evol Comput"}],"container-title":["Journal of Cheminformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s13321-026-01168-8","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s13321-026-01168-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s13321-026-01168-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,22]],"date-time":"2026-04-22T23:05:17Z","timestamp":1776899117000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1186\/s13321-026-01168-8"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,14]]},"references-count":41,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["1168"],"URL":"https:\/\/doi.org\/10.1186\/s13321-026-01168-8","relation":{},"ISSN":["1758-2946"],"issn-type":[{"value":"1758-2946","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,14]]},"assertion":[{"value":"1 December 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"12 February 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 March 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"47"}}