Cited by 440Protein Structure and DynamicsMachine Learning in BioinformaticsEnzyme Structure and Function

The DOI matches its Crossref record; 77 of the 77 references it cites were found in a registry or an index.

OpenFold: retraining AlphaFold2 yields new insights into its learning mechanisms and capacity for generalization

Open access

Ahdritz, Gustaf, Bouatta, Nazim, Floristean, Christina, Kadyan, Sachin, Xia, Qinghui, Gerecke, William, O’Donnell, Timothy J., Berenberg, Daniel, Fisk, Ian, Zanichelli, Niccolò, Zhang, Bo, Nowaczynski, Arkadiusz, Wang, Bei, Stepniewska-Dziubinska, Marta M., Zhang, Shang, Ojewole, Adegoke, Guney, Murat Efe, Biderman, Stella, Watkins, Andrew M., Ra, Stephen, Lorenzo, Pablo Ribalta, Nivon, Lucas, Weitzner, Brian, Ban, Yih-En Andrew, Chen, Shiyang, Zhang, Minjia, Li, Conglong, Song, Shuaiwen Leon, He, Yuxiong, Sorger, Peter K., Mostaque, Emad, Zhang, Zhao, Bonneau, Richard, AlQuraishi, Mohammed·2024·Nature Methods

DOI 10.1038/s41592-024-02272-zPMC PMC11645889

Published version of a preprint

Open this paper in CoChat — free account

Abstract

AlphaFold2 revolutionized structural biology with the ability to predict protein structures with exceptionally high accuracy. Its implementation, however, lacks the code and data required to train new models. These are necessary to (1) tackle new tasks, like protein-ligand complex structure prediction, (2) investigate the process by which the model learns and (3) assess the model's capacity to generalize to unseen regions of fold space. Here we report OpenFold, a fast, memory efficient and trainable implementation of AlphaFold2. We train OpenFold from scratch, matching the accuracy of AlphaFold2. Having established parity, we find that OpenFold is remarkably robust at generalizing even when the size and diversity of its training set is deliberately limited, including near-complete elisions of classes of secondary structure elements. By analyzing intermediate structures produced during training, we also gain insights into the hierarchical manner in which OpenFold learns to fold. In sum, our studies demonstrate the power and utility of OpenFold, which we believe will prove to be a crucial resource for the protein modeling community.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

Integrity signals

Registry

Crossref · 10.1038/s41592-024-02272-z

Cites

77 of 77 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Attention Is All You Need
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 1×
“We used Adam (Kingma and Ba 2015) with 𝛽 1 = 0.9 , 𝛽 2 = 0.99 , and 𝜖 = 10 −6 .”
Influential
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764 · cited 1×
“11) has arguably solved the problem for natural proteins with sufficiently deep MSAs.”
PyTorch: An Imperative Style, High-Performance Deep Learning Library
Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., et al. (2019) · arXiv · doi:10.48550/arxiv.1912.01703 · Cited by 31,473
Dictionary of protein secondary structure: Pattern recognition of hydrogen‐bonded and geometrical features
Kabsch, W., & Sander, C. (1983) · Biopolymers · doi:10.1002/bip.360221211 · Cited by 15,320 · cited 9×
“Regions corresponding to the beta sheets of Lsi1 are predicted by the mainly alpha model with approximately the right shape, except that their atomic coordinates are not sufficiently precise to enable DSSP to classify them as beta sheets.”
MethodInfluential
Comparative Protein Modelling by Satisfaction of Spatial Restraints
Šali, A., & Blundell, T. L. (1993) · Journal of Molecular Biology · doi:10.1006/jmbi.1993.1626 · Cited by 12,977
Scaling Laws for Neural Language Models
Kaplan, J., McCandlish, S., Henighan, T., Brown, T. B., Chess, B., Child, R., Gray, S., Radford, A., Wu, J., & Amodei, D. (2020) · arXiv · doi:10.48550/arxiv.2001.08361 · Cited by 9,550
ColabFold - Making protein folding accessible to all
Mirdita, M., Schütze, K., Moriwaki, Y., Heo, L., Ovchinnikov, S., & Steinegger, M. (2021) · doi:10.1101/2021.08.15.456425 · Cited by 9,391 · cited 1×
“By analyzing intermediate structures produced during training, we also gain insights into the hierarchical manner in which OpenFold learns to fold.”
Physics-informed machine learning
Karniadakis, G. E., Kevrekidis, I. G., Lu, L., Perdikaris, P., Wang, S., & Yang, L. (2021) · Nature Reviews Physics · doi:10.1038/s42254-021-00314-5 · Cited by 8,718 · cited 1×
“For example, integrating physical priors into machine learning models is an area of outstanding scientific interest ([53]).”
AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models
Varadi, M., Anyango, S., Deshpande, M., Nair, S., Natassia, C., Yordanova, G., Yuan, D., Stroe, O., Wood, G., Laydon, A., et al. (2022) · Nucleic Acids Research · doi:10.1093/nar/gkab1061 · Cited by 7,850

Show all 75 references

Sorted by citations

Items 301–400 of 440

Molecular Modeling of the Pathogenetic Mechanisms of Neuropsychiatric Disorders
Abdurazakov, A., Abashkin, D. A., Semina, E. V., Chaika, Y. A., & Golimbet, V. E. (2026) · International Journal of Molecular Sciences · doi:10.3390/ijms27083563 · Cited by 2
Quorum sensing in streptococci and its peptide-mediated modulation
Renshaw, C. P., Rodriguez, K. M., & Tal‐Gan, Y. (2026) · Bioscience Reports · doi:10.1042/bsr20260162 · Cited by 1
Integrating Conformational Sampling and Siamese Learning to Predict Mutation-Induced Binding Affinity Changes in Abelson Tyrosine Kinase and Its Ligands
Xie, L., Ma, T., Lu, X., Zhang, D., Xu, L., Yang, M., Jeong, Y.-S., Chang, S., & Xu, X. (2026) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.6c00695 · Cited by 4
Navigating the uncharted: AI-driven advances in protein structure, dynamics, interactions and ligand interactions for understudied families
Singh, S., Singh, R., & Sharma, S. (2026) · BioData Mining · doi:10.1186/s13040-026-00577-7 · Cited by 2
“Community efforts have further broadened access: OpenFold, a fully open-source and trainable reimplementation of AlphaFold2 from the AlQuraishi laboratory, reproduced AlphaFold2-level accuracy and released its training data, and it remains actively developed [20].”
Influential
Computationally guided strategies to modulate enzyme pH profiles for industrial biocatalysis and environmental applications
Ahmad, M., Iqbal, M. W., Mirza, M. U., Raza, M. A., Alghamdi, H. A., Haider, S. Z., Nawaz, M. Z., & Yu, K. (2026) · Enzyme and Microbial Technology · doi:10.1016/j.enzmictec.2026.110927 · Cited by 1
Adapting Co-Folding Models for Structure-Based Protein-Protein Docking Through Flow Matching
Xu, D. Z., Chu, L., & Gray, J. J. (2026) · PRX Life · doi:10.1103/zfn9-18y7 · Cited by 1
TPS-Flow: Physics-Guided Flow-Based Generative Modeling of Protein Transition Paths
Xu, K., Zhao, L., Tian, Y., Zhou, K., Zang, J., Wang, J., Li, Y., Wang, S., Lin, S., Liu, H., et al. (2026) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6c00807 · Cited by 1
PETIMOT : a novel framework for inferring protein motions from sparse data using SE(3)-equivariant graph neural networks
Lombard, V., Van, J. N., Grudinin, S., & Laine, É. (2026) · Acta Crystallographica Section D Structural Biology · doi:10.1107/s2059798326006054 · Cited by 4
Deep Learning for Protein Modeling: From Single Structure to Bound Complex and Thermodynamic Ensemble, With a Focus on Architectural Design
Angioletti, D., Carli, M., Nobile, M., & Limongelli, V. (2026) · Wiley Interdisciplinary Reviews Computational Molecular Science · doi:10.1002/wcms.70082 · Cited by 1
“…community investigations revealed a key biological insight : the model appears to learn hierarchically, assembling local structural fragments before resolving global folds, a pattern that mirrors physical folding principles and links the abstract training process to protein folding biophysics [38].”
A curated benchmark for cofolding models on kinase conformational states
Sun, K., & Head-Gordon, T. L. (2026) · npj Drug Discovery. · doi:10.1038/s44386-026-00068-z · Cited by 4
Enhancing protein structure prediction accuracy by prioritizing important residues using protein language models
Liu, Y., Kang, B., Li, P., & Cui, Q. (2026) · Communications Biology · doi:10.1038/s42003-026-10725-8 · Cited by 1
How Many Crystal Structures Do You Need to Trust Your Docking Results?
Payne, A. M., Kaminow, B., MacDermott-Opeskin, H., Pulido, I., Scheen, J., Castellanos, M. A., Fearon, D., Chodera, J. D., & Singh, S. (2026) · Journal of Medicinal Chemistry · doi:10.1021/acs.jmedchem.5c03559 · Cited by 1
Assessment of the Performances of the Protein Modeling Techniques Participating in CASP15 Using a Structure-Based Functional Site Prediction Approach: ResiRole
Huang, G. J., Parry, T. K., & McLaughlin, W. A. (2023) · Bioengineering · doi:10.3390/bioengineering10121377 · Cited by 1
“In contrast, OpenFold, which is a fast and trainable implementation of AlphaFold2 [35], uses evolutionary related sequences to build the structure models; but it does not speciﬁcally ﬁlter out sequences that are functionally unrelated.”
Background
An Intelligent Data Analysis System for Biological Macromolecule Crystallography
Sun, H.-K., Hu, Y., Geng, Z. Z., Gao, Z., Zhang, X., & Ding, W. (2024) · EPJ Web of Conferences · doi:10.1051/epjconf/202429505001 · Cited by 1
“The framework provides both automatic selection of optimized parameters across modules and integration of AI algorithms like AlphaFold [18] and OpenFold [19] for predictive modeling.”
Method
MEFold: Memory-Efficient Optimization for Protein Language Models via Chunk and Quantization
Jiang, Y., Sun, N., Lu, Z., Peng, S., Zhang, Y., Yang, F., & Li, T. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10651470 · Cited by 1
“OpenFold [5] reproduces the original AlphaFold2 using PyTorch [19] and enhances performance through various optimization techniques.”
Method
Formation of the Native Topology of a Protein is due to the “Conserved but Non-Functional” Residues: A Case of Apomyoglobin Folding
Bychkova, V. E., Dolgikh, D. A., Balobanov, V. A., & Finkelstein, A. V. (2024) · Frontiers in Bioscience-Landmark · doi:10.31083/j.fbl2911379 · Cited by 1
Successful prediction of LC8 binding to intrinsically disordered proteins sheds light on AlphaFold’s black box
Walker, D. R., Fujimura, G., Vanegas, J. M., & Barbar, E. J. (2025) · Frontiers in Molecular Biosciences · doi:10.3389/fmolb.2025.1531793 · Cited by 1
LightNobel: Improving Sequence Length Limitation in Protein Structure Prediction Model via Adaptive Activation Quantization
Han, S., Choi, S., & Kim, J.-Y. (2025) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1145/3695053.3731006 · Cited by 1
“Recent state-of-the-art PPMs [3, 5, 33, 39] demonstrate exceptional performance through the use of attention mechanisms.”
H3-OPT: Accurate prediction of CDR-H3 loop structures of antibodies with deep learning
Chen, H., Fan, X., Zhu, S., Pei, Y., Zhang, X., Zhang, X., Liu, L., Qian, F., & Tian, B. (2023) · doi:10.1101/2023.08.19.553933 · Cited by 1
Cerebra: A Computationally Efficient Framework for Accurate Protein Structure PredictionPreprint
Gong, H., Hu, J., & Wang, W. (2024) · Research Square · doi:10.21203/rs.3.rs-3995729/v1 · Cited by 1
H3-OPT: Accurate prediction of CDR-H3 loop structures of antibodies with deep learningPreprint
Chen, H., Fan, X., Zhu, S., Pei, Y., Zhang, X., Zhang, X., Liu, L., Qian, F., & Tian, B. (2024) · eLife · doi:10.7554/elife.91512.3 · Cited by 1
Protein Structure Description with ρ, θ and ϕ : A Case Study with Caenopore-5Preprint
Li, W. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.01.08.631847 · Cited by 1
“As the primary source of training and validation data, PDB underpins the performance of most protein structure prediction tools [23,31–33 , 38 ].”
CPI-Pred: A deep learning framework for predicting functional parameters of compound-protein interactionsPreprint
Xu, Z., Barghout, R. A., Wu, J., Garg, D., Song, Y. S., & Mahadevan, R. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.01.16.633372 · Cited by 1
Reinforced molecular dynamics: Physics-infused generative machine learning model explores CRBN activation processPreprint
Kolossváry, I., & Coffey, R. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.02.12.638002 · Cited by 1
Cryptochrome 4b protein is likely irrelevant for the radical pair based magnetoreception in the European robinPreprint
Xu, J., Ameerjan, A. B. P., Hungerland, J., Langebrake, G., Ravnsborg, T., Jensen, O. N., Schmidt, J., Bartölke, R., Kasahara, T., Satish, B., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.02.21.639466 · Cited by 1
“The initial protein structures for both Er Cry4a and Er Cry4b were determined using the unrelaxed predictions from the OpenFold [56] implementation of the AlphaFold [57].”
Harnessing the Evolution of Proteostasis Networks to Reverse Cognitive DysfunctionPreprint
Reineke, L. C., Zhu, P., Dalwadi, U., Dooling, S. W., Liu, Y., Wang, I.-C., Young, S. K., Okoh, J. T., Kuncha, S. K., Zhou, H., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.02.28.640897 · Cited by 1
“OpenFold modeling 49 predicted that DP71L-4X would bulge out and disrupt the salt bridge, assuming a structure similar to ΔPPP1R15B ( Figure 6C ).”
Non-Canonical Crosslinks Confound Evolutionary Protein Structure ModelsPreprint
Lacombe, R. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.03.17.643596 · Cited by 1
“Open-source models provide efficient and accessible alternatives, such as such as ColabFold (Mirdita et al., 2022), OpenFold (Bouatta et al., 2023), and Boltz-1 (Wohlwend et al., 2024).”
LMI4Boltz: Optimising VRAM utilisation to predict large macromolecular complexes with consumer grade hardwarePreprint
Litfin, T., Caley, J. S., & Michie, K. A. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.10.29.684571 · Cited by 1
Constructing the ensemble of representative structures for a protein via neural-surrogate-guided MSA recombinationPreprint
Zhou, H., Yu, H., Yau, S. S. ‐., & Gong, H. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.01.14.699462 · Cited by 1
“After OpenFold evaluation, the anchor entry was replaced if the new structure satisfied: (i) mean pLDDT ≥ δ , (ii) a minimum pLDDT improvement margin p new > p anc + ∆ (we use ∆ = 0 .”
Influential
Blind Challenges Let Us See the Path Forward for Predictive Models
Chodera, J. D., Walters, W. P., Kosuri, S., & Fraser, J. S. (2026) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6c00205 · Cited by 3
GYDE: A collaborative drug discovery platform for AI-powered protein design and engineeringPreprint
Down, T. A., Warowny, M., Walker, A., D’Ascenzo, L., Lee, D., Zhou, Z., Cao, S., Bainbridge, T. W., Nicoludis, J. M., Harris, S. F., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.03.24.714039 · Cited by 1
The Synthetic Epitope Atlas: High-Throughput Design and Validation of De Novo Antibody-Antigen ComplexesPreprint
Altieri, N., Harman, J. L., Noble, D., Murakowska, N. S., Eng, A., McGowan, K. L., Goodnight, D., DiPeso, L., Shikany, C., Engelhart, E., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.04.17.719295 · Cited by 1
Systematic Evaluation of AlphaFold2 and OpenFold3 on Protein–Peptide ComplexesPreprint
Fayetörbay, R., Timuçin, A. C., & Timuçin, E. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.04.21.720042 · Cited by 1
“Open-source implementations such as OpenFold3 (OF3), Boltz-1, Chai-1, and Protenix have expanded access to these architectures [10–12].”
Modelling antibody structures at the speed of languagePreprint
Ellmen, I., Errington, D., Raybould, M. I. J., & Deane, C. M. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.06.03.729879 · Cited by 1
“Fast, memory-efficient methods are used commonly in the pair processing of meth-ods like OpenFold and Boltz [17, 18], however these methods still require quadratic memory and do not fully leverage inner product attention in Invariant Point Attention (IPA).”
HIP-MS: An ultra-high-throughput, sensitive, and versatile affinity enrichment platform for static and dynamic interactome profilingPreprint
Grauvogel, L., Zollbrecht, E., Heymann, T., Brennsteiner, V., Pensl, C., Michaelis, A. C., & Mann, M. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.06.03.729763 · Cited by 1
The first OpenBind release: An open experimental structure–affinity dataset and benchmark for structure-based AIPreprint
Nelen, J., Khan, O., Adams, E. C., Aschenbrenner, J. C., Thompson, W. A., Ebrahim, A., Çapkın, E., Vallée, C., OpenBind, Shotton, E. J., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.08.27.747600 · Cited by 3
“…[2] demonstrated that machine learning (ML)-based approaches can achieve highly accurate protein structure prediction at the proteome scale, while subsequent open-source implementations such as OpenFold [3] also provided training code and data [4], enabling broader model development and adoption.”
Self-citation
GDFold2 : A fast and parallelizable protein folding environment with freely defined objective functions
Mi, T., Xiao, N., & Gong, H. (2025) · Protein Science · doi:10.1002/pro.70041 · Cited by 1
“The detailed implementation could be found in Algorithms 9 – 13 in Data S1.”
Accelerating AlphaFold2 Inference of Protein Three-Dimensional Structure on the Supercomputer Fugaku
Oyama, Y., Tabuchi, A., & Tokuhisa, A. (2023) · doi:10.1145/3589013.3596674 · Cited by 1
“AlphaFold2 is implemented in the JAX deep learning framework, although several other implementations, including its PyTorch version, exist [1, 6].”
MethodInfluential
Probing different conformational states of polar organizing protein Z with AlphaFoldPreprint
Taneja, I., & Lasker, K. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.10.11.561952
H3-OPT: Accurate prediction of CDR-H3 loop structures of antibodies with deep learningPreprint
Chen, H., Fan, X., Zhu, S., Pei, Y., Zhang, X., Zhang, X., Liu, L., Qian, F., & Tian, B. (2023) · eLife · doi:10.7554/elife.91512.1
H3-OPT: Accurate prediction of CDR-H3 loop structures of antibodies with deep learningPreprint
Chen, H., Fan, X., Zhu, S., Pei, Y., Zhang, X., Zhang, X., Liu, L., Qian, F., & Tian, B. (2024) · eLife · doi:10.7554/elife.91512.2
Protein 3D Structure Identification by AlphaFold: a Physics-Based Prediction or Recognition Using Huge Databases?
Finkelstein, A. V. (2024) · The journal of molecular biology. · doi:10.52338/tjomb.2024.3935
“The great success of the AlphaFold, AlphaFold2 and then OpenFold programs [1 3] in identifying three-dimensional (3D) protein structures from their amino acid (a.a.) sequences is obvious [4], but it raises two questions [5]: What is the main reason for this success?”
Background
Assessing the functional roles of coevolving PHD finger residues
Basu, S., Subedi, U., Tonelli, M., Afshinpour, M., Tiwari, N., Fuentes, E. J., & Chakravarty, S. (2024) · Protein Science · doi:10.1002/pro.5065
“MSA and corresponding coevolutionary signals have contributed to the paradigm shift in structure biology (Elofsson, 2023; Jumper & Hassabis, 2022) heralded by AlphaFold2 (Jumper et al., 2021) (AF2), RoseTTAFold (Baek et al., 2021), ESM-2 (Lin et al., 2023), and OpenFold (Ahdritz et al., 2022).”
Background
Protein Design
Ramya, L., Ragunathan, P. T., & Yennamalli, R. M. (2024) · Elsevier eBooks · doi:10.1016/b978-0-323-95502-7.00103-2
AlphaFold-Metainference: Prediction of Structural Ensembles of Disordered ProteinsPreprint
Brotzakis, Z. F., Zhang, S., Murtada, M. H., & Vendruscolo, M. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.11.09.622758 · Cited by 1
Self-citation
Characterising Protein Search Drift using exhaustive protein search and Alphafold2Preprint
Buchan, D. W. A. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.11.14.623594
AlphaPulldown2 – A General Pipeline for High-Throughput Structural Modeling
Molodenskiy, D., Maurer, V. J., Yu, D., Chojnowski, G., Bienert, S., Tauriello, G., Gilep, K., Schwede, T., & Kosinski, J. (2024) · doi:10.1101/2024.11.28.625873
Computational Protein Engineering
Okur, T., & Serçinoğlu, O. (2024) · Elsevier eBooks · doi:10.1016/b978-0-323-95502-7.00261-x
Year in review 2024
(2025) · Nature Methods · doi:10.1038/s41592-024-02591-1
OpenComplex I: Predicting Three-Dimensional Atomic-Level Structures of Biomolecular ComplexesPreprint
Yu, J., Chen, Z., Li, Z., Zeng, M., Lin, W., Huang, H., & Ye, Q. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.03.25.643589 · Cited by 1
“We designed the deep learning architecture based on the layout of OpenFold[Ahdritz et al., 2022], which reimplements the AlphaFold2 architecture consisting of an MSA encoder and an atomic structural decoder (Figure 1).”
GeoFlow-V2: A Unified Atomic Diffusion Model for Protein Structure Prediction and De Novo DesignPreprint
BioGeometry Team (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.05.06.652551 · Cited by 1
“We implement this model using OpenFold’s (Ahdritz et al., 2024) codebase.”
How Risky Are Open Frontier Models?
Krier, S., Law, H., & Klyman, K. (2025) · Oxford University Press eBooks · doi:10.1093/oxfordhb/9780198940272.013.0023
Application of Anomaly Detection to Identify Important Features of Protein Dynamics
Yamamori, Y., & Tomii, K. (2025) · ACS Omega · doi:10.1021/acsomega.4c11546
ClusterEmbed: Efficient Protein Structure Prediction with Clustering and Embeddings
Mu, B. (2025) · BIO Web of Conferences · doi:10.1051/bioconf/202518202013
Over Half a Century of Burial of ρ, θ and ϕ in PDBPreprint
Li, W. (2025) · Qeios · doi:10.32388/o5rm3y
Generative genomics accurately predicts future experimental resultsPreprint
Koytiger, G., Walsh, A. M., Marar, V., Johnson, K. A., Highsmith, M. R., Abbas, A., Stirn, A., Brumbaugh, A. R., David, A., Hui, D., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.09.08.674753 · Cited by 1
“This vision is fast becoming reality for protein structures, where deep learning models excel at not just predicting the structure of existing proteins 1–4 but also designing novel protein structures that do not occur in nature 5–8 .”
Accessible, uniform protein property prediction with a scikit-learn based toolset AIDE
Komp, E., Johansson, K. E., Gauthier, N. P., Gado, J. E., Lindorff‐Larsen, K., & Beckham, G. T. (2025) · Bioinformatics · doi:10.1093/bioinformatics/btaf544 · Cited by 1
“We also provide tools for common tasks on either side of PPP, like predicting structures with Soloseq, creating MSAs with MMseqs2, or sampling designed sequences to maximize predictions with BADASS (Steinegger and S € oding 2017, Ahdritz et al. 2024, Gomez-Uribe et al. 2024).”
MSACLR: Contrastive Learning of Protein Conformations from MSAsPreprint
Zhang, J., Xing, E., & Cheng, X. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.10.01.679854 · Cited by 1
“Each MSA is encoded using the Evoformer module from the OpenFold [25] implementation, with the max msa clusters set to 168 and crop size set to 256.”
Diamond: Harnessing GPU Resources for Scientific Deep Learning
Xie, H., Marwaha, R., Mathew, M., Bian, S., Yang, G., Yan, M., Babuji, Y., Price, O., Wang, Y., Kindratenko, V., et al. (2025) · IEEE International Conference on eScience · doi:10.1109/escience65000.2025.00031
“Our experiences of training OpenFold [8] on NERSC Perl-mutter, TACC Lonestar6, and ALCF Polaris illustrate the challenges.”
Self-citation
Unfolded to Folded: Unraveling the Secrets of Protein Folding with ProteusFoldPreprint
Ahmed, S. S. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.10.08.679099
Explaining Protein Folding Networks Using Integrated Gradients and Attention Mechanisms
Myana, R. S., & Jha, S. K. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-02489-3_15
Frontiers in Multimodal AI: A Survey of Datasets and Benchmarks for Trustworthy Machine Intelligence
Ghajari, G., Majmaie, S. A., Jabo, S. A., Hasan, J., & Amsaad, F. (2025) · IEEE National Aerospace and Electronics Conference (NAECON) · doi:10.1109/naecon65708.2025.11235327
“Efforts such as OpenProteinSet address this gap by providing open, high-quality MSA data for the broader community [5].”
Protein and nucleic acid folding: Domain swapping in proteins
Brosenitsch, T. A., & Gronenborn, A. M. (2025) · Elsevier eBooks · doi:10.1016/b978-0-323-99507-8.00028-7
End-to-end single-stranded DNA sequence design with all-atom structure reconstructionPreprint
Si, Y., Xu, Y., & Chen, L. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2025.12.05.692525
Computational insights into SARS-CoV-2: mutation detection, therapeutic discovery, and vaccine development
Sharmila, A., & Selvaraj, C. I. (2025) · Critical Public Health · doi:10.1080/09581596.2025.2606437
Fine-tuning AlphaFold with limited cryo-EM observations
Liao, J., Zheng, D., Zhang, H., Zhang, L., Hu, M., & Bao, C. (2026) · Communications Chemistry · doi:10.1038/s42004-026-01899-7
“These,aswellasothermethods,havedemonstratedtheusefulness ofdeeplearningtechniquesinthepredictionofproteinstructures 19 – 21 .”
Guiding AlphaFold predictions with experimental knowledge to inform dynamics and interactions with VAIRO
Triviño, J., Jiménez, E., Grininger, C., Caballero, I., Medina, A., Castellví, A., Petrillo, G., Govantes, F., Sagmeister, T., Alcorlo, M., et al. (2026) · Protein Science · doi:10.1002/pro.70481
“Alternative, retrainable implementations such as Open-Fold (Ahdritz et al., 2022) and Uni-Fold (Li et al., 2022) ease integrating other experimental data sources, as in AlphaLink2 (Stahl et al., 2023, 2024), which exploits cross-linking mass spectrometry in the prediction of complexes.”
Transformers as a substrate for structural biology
Malik, A. J., Portelli, S., & Ascher, D. B. (2026) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2025.103218
Disulphide and sequence-encoded conformational priors guide nanobody structure predictionPreprint
Ali, M., Jaskółowski, M., Greenig, M., Nguyen, T. H., Crnogaj, M., Smorodina, E., Ramon, A., Zhao, H., Fernández‐Quintero, M. L., Ghedin, E., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.02.13.705647
“Disulphide bridge loss We adapt the architecture and base training objectives of ABody-Builder3 [72, 73], which in turn are built on the structure module of AlphaFold2.”
Absence of 8-HDF and MTHF Antenna Chromophore Binding in Er CRY4a Suggests a Possible Flavin-Only Cofactor State: Insights from Biochemical and Computational AnalysesPreprint
Ameerjan, A. B. P., Dabirmanesh, B., Hungerland, J., Kasahara, T., Bartölke, R., Dautaj, G., Saberamoli, G., Schmidt, J., Xu, J., Solov’yov, I. A., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.02.17.706321
Prediction of ligand-dependent conformational sampling of ABC transporters by AlphaFold3 and correlation to experimental structures and energeticsPreprint
Tang, Q., Wu, T., Soubasis, B., & Mchaourab, H. S. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.02.19.706864 · Cited by 1
“The emergence of deep learning protein structure prediction methods holds the promise of mapping protein conformational landscape in silico(6-10) .”
Master of Metals2: a graph neural network based architecture for the prediction of zinc binding sites in protein structures
Laveglia, V., Ciofalo, C., Morelli, E., Andreini, C., & Rosato, A. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag078 · Cited by 1
“In the last few years, structure-based tools have b een very popular thanks to the availability of reliable structural models provided by DL methods such as AlphaFold [4], RosettaFold [5] or OpenFold [6].”
Accelerating Lead Optimization Away from Anti-Targets: Navigating the Structural Complexity of Promiscuous ADMET Proteins
Hwang, H., Shelley, M. Y., Placzek, A., DeMarco, K. R., Albanese, S. K., Verras, A., Kaila, N., McElwee, J. J., Campos, S., Caravella, J. A., et al. (2026) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.6c01007
Microbial diversity as a foundation for biological AI
Flamholz, Z., & Kelly, L. (2026) · EMBO Reports · doi:10.1038/s44319-026-00723-6
“A recent paper (Ahdritz et al, 2024) demonstrated that an open, trainable implementation of AlphaFold2 learns structural components hierarchically.”
T-Control: An Efficient Dynamic Tensor Rematerialization System for DNN Training
Wang, Z., Xiao, J., Deng, X., Wang, H., Ma, H., Li, M., Pang, Y., & Tan, G. (2026) · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) · doi:10.1145/3779212.3790230 · Cited by 1
Protein structure prediction
Leach, A. R., & Mobarec, J. C. (2026) · Elsevier eBooks · doi:10.1016/b978-0-443-29808-0.00086-8
Mining antibody functionality via AI-guided structural landscape profiling
Terekhov, S. S., Ivanisenko, N. V., Zhang, N., Mokrushina, Y. A., Nolde, D. E., Lomakin, Y. A., Zalevsky, A. O., Ovchinnikova, L. A., Malabuiok, D. M., Baranova, M. N., et al. (2026) · Nature Communications · doi:10.1038/s41467-026-70553-6
“This structure was further used as an initial template for the AlphaFold2 MSA-free inference 36 (see Supplementary Note).”
Self-citation
Interpretable Machine Learning for Protein Science: Structure, Function, and Interactions
Vecchietti, L. F., Lee, M., Hangeldiyev, B., Wijaya, B. N., Jung, H., Park, H., Kim, T., Cha, M., & Kim, H. M. (2026) · ACM Computing Surveys · doi:10.1145/3801737 · Cited by 5
SurGBSA: learning representations from molecular dynamics simulations
Jones, W. D., Yang, Y., Lightstone, F. C., Moshiri, N., Allen, J. E., & Rosing, T. (2026) · Digital Discovery · doi:10.1039/d6dd00174b
Medicinal chemistry enabling advances in protein structure prediction
Munasinghe, A., Zhang, Z., & Wu, H. (2026) · Elsevier eBooks · doi:10.1016/b978-0-443-29808-0.00096-0
Self-citation
Intelligent Modeling of Protein Function and Drug Discovery Using Large Language Models: A Review
Zhang, M., Liu, B., Xu, J., Zhou, B., Sun, H., Wang, Y., Wang, J., & Xue, M. (2026) · International Journal of Intelligent Systems · doi:10.1155/int/9139921
Self-citation
A geometry-over-coevolution principle governs protein complex assembly in AlphaFoldPreprint
Li, S., Mu, Z., & Yan, C. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.04.03.716280
Structural prediction of protein assemblies with deep learning
Bret, H. (2023) · theses.fr (ABES) · doi:10.70675/b3849091zce01z4ab0z91efz2b71d9cc754d
SimpleFold-Turbo: Adaptive Inference Caching Yields 14-fold Acceleration of Flow-matching Protein Structure PredictionPreprint
Taghon, G. J. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.04.07.714835
“For MSA-reliant architectures such as AlphaFold 3, the MSA computation itself remains a bottleneck ( ≈ 10 s) that caching does not address [11].”
Zero-Shot Generation of Protein Conformational Ensembles Through AlphaFold Latent FloodingPreprint
Qian, R., Zhan, R., Song, Z., & Huang, J. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.04.16.718914
“Here, we introduce AF Latent Flooding (AFLF), a heuristic multiscale framework that repurposes the AF2 model as a zero-shot engine for generating protein conformational ensembles.”
Influential
Ciri: Scaling-aware job scheduling for small language models on over-committed GPU clusters
Wang, J., Fan, S., & Zhang, Z. (2026) · Future Generation Computer Systems · doi:10.1016/j.future.2026.108523
Self-citation
LID-Drug: A Localized Interactive Domain-Aggregated (LID) Framework for Protein Drug Editing
Liu, M., Lou, Y., Luo, S., Yu, Y., Ren, S., & Bai, Y. D. (2026) · Lecture notes in computer science · doi:10.1007/978-981-95-7078-2_12
Machine learning approaches in protein-protein and protein-drug binding
Schlichthaerle, T. (2026) · Elsevier eBooks · doi:10.1016/b978-0-323-99507-8.00061-5
Advances and trends in the bidirectional transformation between biological data and knowledge
Rong, L. I., Ge, J., ZHANG, X.-B., ZHANG, Y.-J., Da-Ming, C. H. E. N., & Tao, C. (2026) · Shengming kexue · doi:10.3724/cbls.2026032
Integrative Structure-Based and AI-Guided Discovery of Natural and Analog UPP1 Inhibitors Targeting the Epigenetic-Metabolic Axis in Lung Adenocarcinoma
De, A. (2026) · Journal of Pharmaceutical Innovation · doi:10.1007/s12247-026-10654-3
“In this study, several state-of-the-art AI-integrated deep learning tools: Boltz-2 [58], Chai-1 [59], IntFold [60], RosettaFold-3 [61], and OpenFold3 [62] were employed to predict the three-dimensional structures of protein–ligand complexes.”
Bayesian-Steered Structure Prediction of Mechanical Biomolecules Using Twisted DiffusionPreprint
Klaus, C., & Sotomayor, M. M. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.05.11.724187
AlphaFold3 predicted LWO G-protein complex from European robin features active-state biased G t αPreprint
Hungerland, J., Kostritski, A., Koch, K., & Solovyov, I. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.05.19.726335
Accurate protein stability prediction for small domains using mega-scale experimentsPreprint
Cho, Y., Tsuboyama, K., Litberg, T. J., Jung, M., Obisesan, A., Wang, Q., Phoumyvong, C. M., Thibeault, J., Ovchinnikov, S., & Rocklin, G. J. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.05.19.726285 · Cited by 6
Generalizable mutation-effect prediction across adaptive immune recognition via unified multimodal framework
Han, R., Zhang, Y., Liu, X., Fu, L., Pan, T., Xu, J., Wang, X. Y., Zhang, P., Chen, X., Lei, J., et al. (2026) · Nature Machine Intelligence · doi:10.1038/s42256-026-01243-7 · Cited by 1
“OpenFold is an open-source reimplementation of AlphaFold2, trained from scratch with publicly Article https://doi.org/10.1038/s42256-026-01243-7 available datasets, and provides competitive accuracy in structure prediction with full transparency and extensibility 68 .”
Influential
MegaFold: Efficient Training of Next-Generation 3D Attention Protein Models on Cross-Platform GPUs
La, H., Gupta, A., Morehead, A., Cheng, J., & Zhang, M. (2026) · ISC High Performance (ISC) · doi:10.23919/isc.2026.11520503 · Cited by 3
“Models such as AlphaFold3 (AF3) extend this capability beyond single proteins to protein-protein, protein-DNA/RNA, and protein-ligand interactions [2], making a shift towards next-generation biomolecular modeling workloads [3]– [5].”
Self-citation
WITHDRAWN: Polyfold: Evaluation of Open-Use Molecular Structure Prediction Algorithms to Inform Their Utility in Diverse Biological ApplicationsPreprint
Stephenson, H., Voicu, D., Novakov, V., Levy, M., & Marsilio, J. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.06.16.732304
Retracted
AbCVista: a deep learning framework for predicting antibody conformational ensembles
Zong, F., Gan, J., Liu, Y., Qian, W., Li, Y., Li, C., Xiao, Z. J., & Cao, Y. (2026) · Protein & Cell · doi:10.1093/procel/pwag040
Self-citation
SteerAF: Distogram-based Steering of AlphaFold2 toward Alternative ConformationsPreprint
Tang, J., Zhu, Z., Yang, S., & Song, C. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.06.19.733296 · Cited by 1
Evaluation of Physics-Based Pose Prediction and the Role of Receptor Flexibility in a Community Antiviral Blind Challenge
Khoualdi, A. F., Clark, F., Horton, J. T., & Cole, D. J. (2026) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6c02582
Editorial: Advances in protein structure biology – use of AI and beyond
Pal, A., Filippidi, E., & Dey, S. (2026) · Frontiers in Molecular Biosciences · doi:10.3389/fmolb.2026.1903318
“Open-source initiatives such as OpenFold are democratizing access to state-of-the-art protein AI models, enabling academic and industrial researchers to adapt and train models on proprietary datasets while maintaining data privacy (Ahdritz et al., 2024).”

Authors: Ahdritz, G., Bouatta, N., Floristean, C., Kadyan, S., Xia, Q., Gerecke, W., O’Donnell, T. J., Berenberg, D., Fisk, I., Zanichelli, N., Zhang, B., Nowaczynski, A., Wang, B., Stepniewska-Dziubinska, M. M., Zhang, S., Ojewole, A., Guney, M. E., Biderman, S., Watkins, A. M., Ra, S., Lorenzo, P. R., Nivon, L., Weitzner, B., Ban, Y.-E. A., Chen, S., Zhang, M., Li, C., Song, S. L., He, Y., Sorger, P. K., Mostaque, E., Zhang, Z., Bonneau, R., & AlQuraishi, M.
Affiliations: Harvard University (US)
Columbia University (US)
Icahn School of Medicine at Mount Sinai (US)
Courant Institute of Mathematical Sciences (US)
Flatiron Institute (US)
University of Utah (US)
Nvidia (United States) (US)
Booz Allen Hamilton (United States) (US)
Genentech (US)
Arzeda (United States) (US)
Rutgers, The State University of New Jersey (US)
University of Illinois Urbana-Champaign (US)
Microsoft (United States) (US)
Corresponding author: Bouatta, Nazim
AlQuraishi, Mohammed
ORCID: Ahdritz, Gustaf
Bouatta, Nazim
Floristean, Christina
Kadyan, Sachin
Xia, Qinghui
Gerecke, William
O’Donnell, Timothy J.
Berenberg, Daniel
Fisk, Ian
Zanichelli, Niccolò
Zhang, Bo
Nowaczynski, Arkadiusz
Wang, Bei
Stepniewska-Dziubinska, Marta M.
Zhang, Shang
Ojewole, Adegoke
Biderman, Stella
Watkins, Andrew M.
Ra, Stephen
Lorenzo, Pablo Ribalta
Weitzner, Brian
Ban, Yih-En Andrew
Chen, Shiyang
Zhang, Minjia
Song, Shuaiwen Leon
He, Yuxiong
Sorger, Peter K.
Zhang, Zhao
Bonneau, Richard
AlQuraishi, Mohammed
Venue: Nature Methods
Publisher: Springer Science and Business Media LLC
Published: 2024-05-14
Volume: 21
Issue: 8
Pages: 1514-1524
ISSN: 1548-7091 (ISSN-L)
1548-7105
DOI: 10.1038/s41592-024-02272-z
PMC: PMC11645889
OpenAlex: W4396895148
Cited by: 439
References: 75
Citations by year: 2026: 133 · 2025: 138 · 2024: 107
FWCI: 51.31
Type: journal article
Language: en
Open access: hybrid
doi.org
Licence: cc-by
www.springernature.com/gp/researchers/text-and-data-mining (tdm)
www.springernature.com/gp/researchers/text-and-data-mining (vor)
Locations: Nature Methods (publishedVersion, cc-by)
bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
PubMed (publishedVersion)
PubMed Central (acceptedVersion, cc-by)
Digital Access to Scholarship at Harvard (DASH) (Harvard University) (acceptedVersion)
Funders: U.S. Department of Health & Human Services | NIH | National Institute of General Medical Sciences (R35GM150546)
U.S. Department of Health & Human Services | NIH | National Cancer Institute (U54-CA225088)
National Science Foundation (OAC-2106661, OAC-2112606)
Topics: Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
Life Sciences › Biochemistry, Genetics and Molecular Biology › Structural Biology › Enzyme Structure and Function
Keywords: AlphaFold2, protein structure prediction, protein folding, model generalization, secondary structure elements, learning mechanisms, hierarchical folding
MeSH terms: Algorithms, Models, Molecular, Protein Conformation, Proteins, Software, Protein Structure, Secondary, Protein Folding, Computational Biology

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

OpenFold: retraining AlphaFold2 yields new insights into its learning mechanisms and capacity for generalization

Abstract

Overview

Integrity signals

Verification

Claims the paper makes

Claims other papers make about this one

Self-check

Citations

Cites

Cited by

Related

Metadata

Record

About these checks