{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T21:19:23Z","timestamp":1792790363517,"version":"4.5.0"},"posted":{"date-parts":[[2026,9,18]]},"group-title":"Bioinformatics","reference-count":95,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T00:00:00Z","timestamp":1789689600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"award":["2139291"],"award-info":[{"award-number":["2139291"]}],"id":[{"id":"https:\/\/ror.org\/021nxhr62","id-type":"ROR","asserted-by":"publisher"}]},{"award":["T32 GM115327"],"award-info":[{"award-number":["T32 GM115327"]}],"id":[{"id":"https:\/\/ror.org\/01cwqze88","id-type":"ROR","asserted-by":"publisher"}]},{"award":["R35 GM152017"],"award-info":[{"award-number":["R35 GM152017"]}],"id":[{"id":"https:\/\/ror.org\/01cwqze88","id-type":"ROR","asserted-by":"publisher"}]},{"award":["K99 CA286801"],"award-info":[{"award-number":["K99 CA286801"]}],"id":[{"id":"https:\/\/ror.org\/040gcmg81","id-type":"ROR","asserted-by":"publisher"}]},{"award":["P30 CA008748"],"award-info":[{"award-number":["P30 CA008748"]}],"id":[{"id":"https:\/\/ror.org\/040gcmg81","id-type":"ROR","asserted-by":"publisher"}]},{"name":"Damon Runyon Cancer Research Foundation","award":["DRQ-14-22"],"award-info":[{"award-number":["DRQ-14-22"]}],"id":[{"id":"https:\/\/ror.org\/01gd7b947","id-type":"ROR","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,9,18]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>\n                  Structure-based drug discovery is a widely used paradigm for the rational design of novel small molecule therapeutics. However, the benefits conferred by the use of structural information has seen limited adoption in machine learning, where ligand-only (\"2D\") models are still the industry standard for molecular property or binding affinity prediction. Structure-based (\"3D\") ML models for binding-affinity prediction promise to present a clear advantage, but have not yet overtaken existing 2D models. Here, we show that physics-based priors can improve predictive performance of structure-based models by comparing different model architectures with varying physical priors on several prediction tasks. We present the Modular Training and Evaluation of Neural Networks (\n                  <jats:monospace>mtenn<\/jats:monospace>\n                  ) package, where we decompose affinity prediction into separate steps of embedding structure into learned representations and combining those embeddings into a predicted binding affinity. We consider both E(3)-invariant and E(3)-equivariant architectures to determine the importance of encoding roto-translational inductive biases, as well as different methods for combining learned embeddings. By first optimizing several aspects of model construction using the general purpose PDBBind dataset, we are able to improve the performance and data efficiency of structure-based models. When subsequently trained and evaluated on the COVID Moonshot small molecule drug discovery dataset, our tuned models perform on par with industry standard ligand-only models. Our decomposed model framework highlights that encoding some physical priors improves model performance, while more complex biases such as equivariance offer limited benefit. Additionally, structure-based models generalize better to an unseen target and display higher training efficiency. Overall, these results emphasize that structure-based models benefit from their ability to incorporate physics-informed constraints, giving promising directions for model architecture development. These results also suggest that the strength of these models may be in tasks specifically aimed at generalizability, providing guidelines for their use in early-stage drug discovery campaigns.\n                <\/jats:p>","DOI":"10.64898\/2026.09.11.750982","type":"posted-content","created":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T19:50:18Z","timestamp":1789761018000},"source":"Crossref","is-referenced-by-count":0,"title":["Physical priors improve performance of structure-based binding affinity models"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-2266-3353","authenticated-orcid":false,"given":"Benjamin","family":"Kaminow","sequence":"first","affiliation":[{"name":"Tri-Institutional Ph.D. Program in Computational Biology & Medicine, Weill Cornell Medical College, New York, New York 10065, United States"},{"name":"Computational and Systems Biology Program, Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, New York, N.Y. 10065, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0947-0191","authenticated-orcid":false,"given":"Alexander Matthew","family":"Payne","sequence":"additional","affiliation":[{"name":"Computational and Systems Biology Program, Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, New York, N.Y. 10065, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7393-7457","authenticated-orcid":false,"given":"Hugo","family":"MacDermott-Opeskin","sequence":"additional","affiliation":[{"name":"Open Molecular Software Foundation, Davis CA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0542-119X","authenticated-orcid":false,"given":"John D.","family":"Chodera","sequence":"additional","affiliation":[{"name":"Computational and Systems Biology Program, Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, New York, N.Y. 10065, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1914-4955","authenticated-orcid":false,"given":"Sukrit","family":"Singh","sequence":"additional","affiliation":[{"name":"Computational and Systems Biology Program, Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, New York, N.Y. 10065, United States"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026092213000766000_2026.09.11.750982v1.1","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.5b00142"},{"key":"2026092213000766000_2026.09.11.750982v1.2","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuron.2018.08.011"},{"key":"2026092213000766000_2026.09.11.750982v1.3","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jmedchem.3c00521"},{"key":"2026092213000766000_2026.09.11.750982v1.4","doi-asserted-by":"publisher","DOI":"10.1007\/s10822-011-9518-x"},{"key":"2026092213000766000_2026.09.11.750982v1.5","doi-asserted-by":"publisher","DOI":"10.1016\/j.sbi.2026.103257"},{"key":"2026092213000766000_2026.09.11.750982v1.6","doi-asserted-by":"publisher","DOI":"10.1038\/s41388-024-03077-2"},{"key":"2026092213000766000_2026.09.11.750982v1.7","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-63602-z"},{"key":"2026092213000766000_2026.09.11.750982v1.8","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-019-0917-9"},{"issue":"10","key":"2026092213000766000_2026.09.11.750982v1.9","doi-asserted-by":"crossref","first-page":"4799","DOI":"10.1038\/s41596-021-00597-z","article-title":"A practical guide to large-scale docking","volume":"16","year":"2021","journal-title":"Nature protocols"},{"key":"2026092213000766000_2026.09.11.750982v1.10","unstructured":"Alexander Matthew Payne , Benjamin Kaminow , Hugo MacDermott-Opeskin , Iv\u00e1n Pulido , Jenke Scheen , Maria A. Castellanos , Daren Fearon , John D. Chodera , and Sukrit Singh . How many crystal structures do you need to trust your docking results?, September 2025. URL https:\/\/www.biorxiv.org\/content\/10.1101\/2025.09.19.677428v1. ISSN: 2692-8205 Pages: 2025.09.19.677428 Section: New Results."},{"key":"2026092213000766000_2026.09.11.750982v1.11","doi-asserted-by":"publisher","DOI":"10.1021\/ja512751q"},{"issue":"1","key":"2026092213000766000_2026.09.11.750982v1.12","first-page":"18378","article-title":"Best practices for alchemical free energy calculations [article v1. 0]","volume":"2","year":"2020","journal-title":"Living journal of computational molecular science"},{"key":"2026092213000766000_2026.09.11.750982v1.13","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.7b00650"},{"key":"2026092213000766000_2026.09.11.750982v1.14","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btu626"},{"key":"2026092213000766000_2026.09.11.750982v1.15","doi-asserted-by":"publisher","DOI":"10.1016\/j.sbi.2023.102548"},{"key":"2026092213000766000_2026.09.11.750982v1.16","doi-asserted-by":"publisher","DOI":"10.1186\/s13321-023-00795-9"},{"key":"2026092213000766000_2026.09.11.750982v1.17","doi-asserted-by":"crossref","unstructured":"Saro Passaro , Gabriele Corso , Jeremy Wohlwend , Mateo Reveiz , Stephan Thaler , Vignesh Ram Somnath , Noah Getz , Tally Portnoi , Julien Roy , Hannes Stark , David Kwabi-Addo , Dominique Beaini , Tommi Jaakkola , and Regina Barzilay . Boltz-2: Towards Accurate and Efficient Binding Affinity Prediction, June 2025. URL https:\/\/www.biorxiv.org\/content\/10.1101\/2025.06.14.659707v1. Pages: 2025.06.14.659707 Section: New Results.","DOI":"10.1101\/2025.06.14.659707"},{"key":"2026092213000766000_2026.09.11.750982v1.18","doi-asserted-by":"publisher","DOI":"10.1002\/advs.202405404"},{"key":"2026092213000766000_2026.09.11.750982v1.19","doi-asserted-by":"publisher","DOI":"10.1039\/c9sc04606b"},{"key":"2026092213000766000_2026.09.11.750982v1.20","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btq112"},{"key":"2026092213000766000_2026.09.11.750982v1.21","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.5c02106"},{"key":"2026092213000766000_2026.09.11.750982v1.22","doi-asserted-by":"publisher","DOI":"10.1016\/j.bpj.2023.11.2596"},{"key":"2026092213000766000_2026.09.11.750982v1.23","unstructured":"[23] Limei Wang , Haoran Liu , Yi Liu , Jerry Kurtin , and Shuiwang Ji . Learning Hierarchical Protein Representations via Complete 3D Graph Networks, March 2023. URL http:\/\/arxiv.org\/abs\/2207.12600. arXiv:2207.12600 [cs.LG]."},{"key":"2026092213000766000_2026.09.11.750982v1.24","doi-asserted-by":"publisher","DOI":"10.1039\/d3sc04185a"},{"key":"2026092213000766000_2026.09.11.750982v1.25","doi-asserted-by":"publisher","DOI":"10.1038\/s41594-026-01797-5"},{"key":"2026092213000766000_2026.09.11.750982v1.26","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.2c01149"},{"key":"2026092213000766000_2026.09.11.750982v1.27","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jmedchem.2c00487"},{"key":"2026092213000766000_2026.09.11.750982v1.28","doi-asserted-by":"crossref","unstructured":"Bj\u00f6rn Mattsson and W Patrick Walters . Identifying and addressing systematic data leakage in protein-ligand affinity benchmarks. bioRxiv, pages 2026\u201306, 2026.","DOI":"10.64898\/2026.06.29.735309"},{"key":"2026092213000766000_2026.09.11.750982v1.29","unstructured":"Openbind: Powering the future of drug discovery with ai. URL http:\/\/openbind.uk."},{"key":"2026092213000766000_2026.09.11.750982v1.30","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-025-01124-5"},{"key":"2026092213000766000_2026.09.11.750982v1.31","doi-asserted-by":"publisher","DOI":"10.1039\/D0CS00098A"},{"key":"2026092213000766000_2026.09.11.750982v1.32","doi-asserted-by":"publisher","DOI":"10.1021\/jm4004285"},{"key":"2026092213000766000_2026.09.11.750982v1.33","unstructured":"Benoit Beck , Yun-Fei Chen , Walthere Dere , Viswanath Devanarayan , Brian J. Eastwood , Mark W. Farmen , Stephen J. Iturria , Phillip W. Iversen , Steven D. Kahl , Roger A. Moore , Barry D. Sawyer , and Jeffrey Weidner . Assay Operations for SAR Support. In Sarine Markossian , Abigail Grossman , Hannah Baskir , Michelle Arkin , Douglas Auld , Christopher Austin , Jonathan Baell , Kyle Brimacombe , Thomas D. Y. Chung , Nathan P. Coussens , Jayme L. Dahlin , Viswanath Devanarayan , Timothy L. Foley , Marcie Glicksman , Kirill Gorshkov , Stefan Grotegut , Matthew D. Hall , Samuel Hoare , James Inglese , Philip W. Iversen , Madhu Lal-Nag , Zhuyin Li , Jason R. Manro , James McGee , Allison Norvil , Mackenzie Pearson , Terry Riss , Peter Saradjian , G. Sitta Sittampalam , Michael A. Tarselli , O. Joseph Trask , Jeffrey R. Weidner , Mary Jo Wildey , Kelli Wilson , Menghang Xia , and Xin Xu , editors, Assay Guidance Manual. Eli Lilly & Company and the National Center for Advancing Translational Sciences, Bethesda (MD), 2004. URL http:\/\/www.ncbi.nlm.nih.gov\/books\/NBK91994\/."},{"key":"2026092213000766000_2026.09.11.750982v1.34","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btz665"},{"key":"2026092213000766000_2026.09.11.750982v1.35","doi-asserted-by":"publisher","DOI":"10.1021\/acscentsci.8b00507"},{"key":"2026092213000766000_2026.09.11.750982v1.36","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0220113"},{"key":"2026092213000766000_2026.09.11.750982v1.37","doi-asserted-by":"crossref","unstructured":"Saro Passaro , Gabriele Corso , Jeremy Wohlwend , Mateo Reveiz , Stephan Thaler , Vignesh Ram Somnath , Noah Getz , Tally Portnoi , Julien Roy , Hannes Stark , David Kwabi-Addo , Dominique Beaini , Tommi Jaakkola , and Regina Barzilay . Boltz-2: Towards Accurate and Efficient Binding Affinity Prediction, June 2025. URL https:\/\/www.biorxiv.org\/content\/10.1101\/2025.06.14.659707v1. Pages: 2025.06.14.659707 Section: New Results.","DOI":"10.1101\/2025.06.14.659707"},{"key":"2026092213000766000_2026.09.11.750982v1.38","unstructured":"Shunzhou Wan , Xibei Zhang , Xiao Xue , and Peter V. Coveney . On the Reliability of AI Methods in Drug Discovery: Evaluation of Boltz-2 for Structure and Binding Affinity Prediction, 2026. URL https:\/\/arxiv.org\/abs\/2603.05532. Version Number: 1."},{"key":"2026092213000766000_2026.09.11.750982v1.39","unstructured":"Shunzhou Wan , Xibei Zhang , Xiao Xue , and Peter V. Coveney . On the Reliability of AI Methods in Drug Discovery: Evaluation of Boltz-2 for Structure and Binding Affinity Prediction, March 2026. URL http:\/\/arxiv.org\/abs\/2603.05532. arXiv:2603.05532 [physics.chem-ph]."},{"key":"2026092213000766000_2026.09.11.750982v1.40","unstructured":"Ryan. Evaluating Boltz-2 on Real Drug Targets: Does it work? URL https:\/\/www.deepmirror.ai\/post\/boltz-2-real-drug-targets."},{"key":"2026092213000766000_2026.09.11.750982v1.41","doi-asserted-by":"crossref","unstructured":"Malte Grieswelle , Samuel Homberg , Philipp Janssen , Johannes Kaminski , Joana Massa , Ann-Kathrin Prinz , Stefan Georgiev , and Oliver Koch . A new benchmark for deep learning based affinity prediction: Solving the inter-protein scoring noise problem, January 2026. URL https:\/\/chemrxiv.org\/doi\/full\/10.26434\/chemrxiv-2025-sf3cs\/v2.","DOI":"10.26434\/chemrxiv-2025-sf3cs\/v2"},{"key":"2026092213000766000_2026.09.11.750982v1.42","doi-asserted-by":"crossref","unstructured":"Maria A. Castellanos , Alexander M. Payne , Jenke Scheen , Hugo MacDermott-Opeskin , Iv\u00e1n Pulido , Blake H. Balcomb , Ed J. Griffen , Daren Fearon , Haim Barr , Noa Lahav , David Cousins , Jessica Stacey , Ralph Robinson , Bruce Lefker , and John D. Chodera . A Structure-Based Computational Pipeline for Broad-Spectrum Antiviral Discovery, July 2025. URL https:\/\/www.biorxiv.org\/content\/10.1101\/2025.07.29.667267v1. ISSN: 2692-8205 Pages: 2025.07.29.667267 Section: New Results.","DOI":"10.1101\/2025.07.29.667267"},{"key":"2026092213000766000_2026.09.11.750982v1.43","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-022-29939-5"},{"key":"2026092213000766000_2026.09.11.750982v1.44","doi-asserted-by":"publisher","DOI":"10.1007\/s10822-016-9938-8"},{"key":"2026092213000766000_2026.09.11.750982v1.45","unstructured":"Justin Gilmer , Samuel S. Schoenholz , Patrick F. Riley , Oriol Vinyals , and George E. Dahl . Neural Message Passing for Quantum Chemistry, June 2017. URL http:\/\/arxiv.org\/abs\/1704.01212. arXiv:1704.01212 [cs.LG]."},{"key":"2026092213000766000_2026.09.11.750982v1.46","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.9b00237"},{"key":"2026092213000766000_2026.09.11.750982v1.47","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.6b00740"},{"key":"2026092213000766000_2026.09.11.750982v1.48","unstructured":"Kristof T. Sch\u00fctt , Pieter-Jan Kindermans , Huziel E. Sauceda , Stefan Chmiela , Alexandre Tkatchenko , and Klaus-Robert M\u00fcller . SchNet: A continuous-filter convolutional neural network for modeling quantum interactions, December 2017. URL http:\/\/arxiv.org\/abs\/1706.08566. arXiv:1706.08566 [stat.ML]."},{"key":"2026092213000766000_2026.09.11.750982v1.49","unstructured":"Bowen Jing , Stephan Eismann , Patricia Suriana , Raphael J. L. Townshend , and Ron Dror . Learning from Protein Structure with Geometric Vector Perceptrons, May 2021. URL http:\/\/arxiv.org\/abs\/2009.01411. arXiv:2009.01411 [q-bio.BM]."},{"key":"2026092213000766000_2026.09.11.750982v1.50","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.6459381"},{"key":"2026092213000766000_2026.09.11.750982v1.51","unstructured":"Taco S. Cohen and Max Welling . Group Equivariant Convolutional Networks, June 2016. URL http:\/\/arxiv.org\/abs\/1602.07576. arXiv:1602.07576 [cs]."},{"key":"2026092213000766000_2026.09.11.750982v1.52","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-97442-7_3"},{"key":"2026092213000766000_2026.09.11.750982v1.53","unstructured":"Maurice Weiler , Mario Geiger , Max Welling , Wouter Boomsma , and Taco Cohen . 3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data, 2018. URL https:\/\/arxiv.org\/abs\/1807.02547. _eprint: 1807.02547."},{"key":"2026092213000766000_2026.09.11.750982v1.54","doi-asserted-by":"publisher","DOI":"10.1126\/science.abo7201"},{"key":"2026092213000766000_2026.09.11.750982v1.55","doi-asserted-by":"publisher","DOI":"10.1186\/s13321-020-00460-5"},{"key":"2026092213000766000_2026.09.11.750982v1.56","doi-asserted-by":"publisher","DOI":"10.1021\/ci9800211"},{"key":"2026092213000766000_2026.09.11.750982v1.57","doi-asserted-by":"publisher","DOI":"10.1016\/j.procs.2024.09.243"},{"key":"2026092213000766000_2026.09.11.750982v1.58","unstructured":"Zhichun Guo , Kehan Guo , Bozhao Nan , Yijun Tian , Roshni G Iyer , Yihong Ma , Olaf Wiest , Xiangliang Zhang , Wei Wang , Chuxu Zhang , and Nitesh V Chawla . Graph-based Molecular Representation Learning."},{"key":"2026092213000766000_2026.09.11.750982v1.59","doi-asserted-by":"publisher","DOI":"10.1186\/s13321-023-00795-9"},{"key":"2026092213000766000_2026.09.11.750982v1.60","doi-asserted-by":"publisher","DOI":"10.1186\/s13321-021-00506-2"},{"key":"2026092213000766000_2026.09.11.750982v1.61","first-page":"18822","volume-title":"Advances in Neural Information Processing Systems","volume":"35","year":"2022"},{"key":"2026092213000766000_2026.09.11.750982v1.62","unstructured":"Guolin Ke , Qi Meng , Thomas Finley , Taifeng Wang , Wei Chen , Weidong Ma , Qiwei Ye , and Tie-Yan Liu . LightGBM: A Highly Efficient Gradient Boosting Decision Tree. In Advances in Neural Information Processing Systems, volume 30. Curran Associates, Inc., 2017. URL https:\/\/proceedings.neurips.cc\/paper\/2017\/hash\/6449f44a102fde848669bdd9eb6b76fa-Abstract.html."},{"key":"2026092213000766000_2026.09.11.750982v1.63","unstructured":"Why Don\u2019t Machine Learning Models Extrapolate?, April 2025. URL https:\/\/patwalters.github.io\/Why-Dont-Machine-Learning-Models-Extrapolate\/."},{"key":"2026092213000766000_2026.09.11.750982v1.64","doi-asserted-by":"publisher","DOI":"10.1186\/s13321-020-00479-8"},{"key":"2026092213000766000_2026.09.11.750982v1.65","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1509.09292"},{"key":"2026092213000766000_2026.09.11.750982v1.66","unstructured":"T. Konstantin Rusch , Michael M. Bronstein , and Siddhartha Mishra . A Survey on Oversmoothing in Graph Neural Networks, March 2023. URL http:\/\/arxiv.org\/abs\/2303.10993. arXiv:2303.10993 [cs]."},{"key":"2026092213000766000_2026.09.11.750982v1.67","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2502.10818"},{"key":"2026092213000766000_2026.09.11.750982v1.68","unstructured":"Mitchell Black , Zhengchao Wan , Amir Nayyeri , and Yusu Wang . Understanding Oversquashing in GNNs through the Lens of Effective Resistance, June 2023. URL http:\/\/arxiv.org\/abs\/2302.06835. arXiv:2302.06835 [cs.LG]."},{"key":"2026092213000766000_2026.09.11.750982v1.69","unstructured":"Shaked Brody , Uri Alon , and Eran Yahav . How Attentive are Graph Attention Networks?, 2021. URL https:\/\/arxiv.org\/abs\/2105.14491."},{"key":"2026092213000766000_2026.09.11.750982v1.70","doi-asserted-by":"publisher","DOI":"10.1016\/j.compchemeng.2023.108202"},{"key":"2026092213000766000_2026.09.11.750982v1.71","unstructured":"Gabriele Corso , Luca Cavalleri , Dominique Beaini , Pietro Li\u00f2 , and Petar Veli\u010dkovi\u0107 . Principal Neighbourhood Aggregation for Graph Nets, December 2020. URL http:\/\/arxiv.org\/abs\/2004.05718. arXiv:2004.05718 [cs.LG]."},{"key":"2026092213000766000_2026.09.11.750982v1.72","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.4c00104"},{"key":"2026092213000766000_2026.09.11.750982v1.73","unstructured":"Xuefeng Liu , Songhao Jiang , Xiaotian Duan , Archit Vasan , Chong Liu , Chih-chan Tien , Heng Ma , Thomas Brettin , Fangfang Xia , Ian T. Foster , and Rick L. Stevens . Binding Affinity Prediction: From Conventional to Machine Learning-Based Approaches, September 2024. URL http:\/\/arxiv.org\/abs\/2410.00709. arXiv:2410.00709 [q-bio] version: 1."},{"key":"2026092213000766000_2026.09.11.750982v1.74","unstructured":"Mario Geiger and Tess Smidt . e3nn: Euclidean Neural Networks, July 2022. URL http:\/\/arxiv.org\/abs\/2207.09453. arXiv:2207.09453 [cs]."},{"key":"2026092213000766000_2026.09.11.750982v1.75","unstructured":"Risi Kondor , Zhen Lin , and Shubhendu Trivedi . Clebsch-Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Network, 2018. URL https:\/\/arxiv.org\/abs\/1806.09231. _eprint: 1806.09231."},{"key":"2026092213000766000_2026.09.11.750982v1.76","unstructured":"Nathaniel Thomas , Tess Smidt , Steven Kearnes , Lusann Yang , Li Li , Kai Kohlhoff , and Patrick Riley . Tensor field networks: Rotation-and translation-equivariant neural networks for 3D point clouds, 2018. URL https:\/\/arxiv.org\/abs\/1802.08219. _eprint: 1802.08219."},{"key":"2026092213000766000_2026.09.11.750982v1.77","doi-asserted-by":"publisher","DOI":"10.1016\/0006-2952(73)90196-2"},{"key":"2026092213000766000_2026.09.11.750982v1.78","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.8b00545"},{"key":"2026092213000766000_2026.09.11.750982v1.79","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0061007"},{"key":"2026092213000766000_2026.09.11.750982v1.80","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.4c00049"},{"key":"2026092213000766000_2026.09.11.750982v1.81","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jpcb.5c08598"},{"key":"2026092213000766000_2026.09.11.750982v1.82","doi-asserted-by":"crossref","unstructured":"Bj\u00f6rn Mattsson and W. Patrick Walters . Identifying and Addressing Systematic Data Leakage in Protein-Ligand Affinity Benchmarks, June 2026. URL https:\/\/www.biorxiv.org\/content\/10.64898\/2026.06.29.735309v1. ISSN: 2692-8205 Pages: 2026.06.29.735309 Section: New Results.","DOI":"10.64898\/2026.06.29.735309"},{"key":"2026092213000766000_2026.09.11.750982v1.83","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.2c01149"},{"key":"2026092213000766000_2026.09.11.750982v1.84","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.2c00785"},{"key":"2026092213000766000_2026.09.11.750982v1.85","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.5c01609"},{"key":"2026092213000766000_2026.09.11.750982v1.86","doi-asserted-by":"publisher","DOI":"10.1021\/ci400084k"},{"key":"2026092213000766000_2026.09.11.750982v1.87","doi-asserted-by":"publisher","DOI":"10.1021\/ci100031x"},{"key":"2026092213000766000_2026.09.11.750982v1.88","doi-asserted-by":"publisher","DOI":"10.1021\/acsomega.1c04017"},{"key":"2026092213000766000_2026.09.11.750982v1.89","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pcbi.1011435"},{"key":"2026092213000766000_2026.09.11.750982v1.90","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jcim.3c00799"},{"key":"2026092213000766000_2026.09.11.750982v1.91","unstructured":"OpenBind-0: Advancing Open Molecular Structure Prediction. URL https:\/\/openbind.uk\/news\/blog-openbind-0-advancing-open-molecular-structure-prediction\/."},{"key":"2026092213000766000_2026.09.11.750982v1.92","doi-asserted-by":"crossref","unstructured":"Jochem Nelen , Omeir Khan , Etowah Adams , Jasmin C. Aschenbrenner , Warren Thompson , Ali Ebrahim , Eda \u00c7apkin , C\u00e9dric Vall\u00e9e , OpenBind, Elizabeth J. Shotton , Ed J. Griffen , John D. Chodera , Charlotte M. Deane , Frank Von Delft , Mohammed AlQuraishi , and Fergus Imrie . The first OpenBind release: An open experimental structure\u2013affinity dataset and benchmark for structure-based AI, August 2026. URL http:\/\/biorxiv.org\/lookup\/doi\/10.64898\/2026.08.27.747600.","DOI":"10.64898\/2026.08.27.747600"},{"key":"2026092213000766000_2026.09.11.750982v1.93","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jpcb.5c08598"},{"key":"2026092213000766000_2026.09.11.750982v1.94","unstructured":"XChem @ Diamond. URL https:\/\/fragalysis.diamond.ac.uk\/viewer\/react\/landing."},{"key":"2026092213000766000_2026.09.11.750982v1.95","doi-asserted-by":"publisher","DOI":"10.1016\/j.econlet.2023.111369"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.09.11.750982","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T20:00:25Z","timestamp":1790107225000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.64898\/2026.09.11.750982"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,18]]},"references-count":95,"URL":"https:\/\/doi.org\/10.64898\/2026.09.11.750982","relation":{},"subject":[],"published":{"date-parts":[[2026,9,18]]},"subtype":"preprint"}}