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 accessPublished version of a preprint
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.
Overview
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
Cites
Statistics
Verification
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“allow users to run a modified AF2 protocol on open access servers in minutes”
From The power and pitfalls of AlphaFold2 for structure prediction beyond rigid globular proteins (2024) · cites this paper as [9]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
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 .”
InfluentialHighly 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
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.”
MethodInfluentialComparative 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]).”
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
Cited by
Sorted by citations
Items 201–300 of 440
Controllable All-Atom Protein Generation with Latent DiffusionPreprint
Lu, A. X., Yang, J., Long, Y., Lambert, T., Yan, W., Robinson, S. A., Yang, K., Gligorijević, V., Cho, K., Bonneau, R., et al. (2024) · Research Square · doi:10.1101/2024.12.02.626353 · Cited by 6
Self-citationThakur, Z., Bansal, L., & Mehta, P. K. (2025) · Future Medicinal Chemistry · doi:10.1080/17568919.2025.2608284 · Cited by 5
ProkBERT PhaStyle: accurate phage lifestyle prediction with pretrained genomic language models
Juhász, J., Ligeti-Nagy, N., Bodnár, B., Juhász, J., Pongor, S., & Ligeti, B. (2024) · Bioinformatics Advances · doi:10.1093/bioadv/vbaf188 · Cited by 4
“In biological sequences, transformer models such as MSA Transformer [22], ESM [23, 24], AlphaFold2 [25], and openFold [26] have been effectively applied.”
Conformation-aware structure prediction of antigen-recognizing immune proteins
Dreyer, F. A., Ludwiczak, J., Martinkus, K., Abanades, B., Alberstein, R. G., Kessel, P., Rao, P., Lee, J. H., Bonneau, R., Watkins, A. M., et al. (2025) · mAbs · doi:10.1080/19420862.2025.2602217 · Cited by 8
“The model implementation leverages code from OpenFold 53 and ABodyBuilder3.”
Self-citationTaneja, I., Llanos, M. A., Fernández‐Quintero, M. L., Loeffler, J. R., Holcomb, M., Ward, A. B., & Forli, S. (2025) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.5c02004 · Cited by 4
Beyond sequence similarity: toward function-based screening of nucleic acid synthesis
Abel, G. R., Alexanian, T., Bartling, C., Beal, J., Curtis, S., Flyangolts, K., Foner, L. N., Forry, S. P., Godbold, G. D., Horvitz, E., et al. (2026) · Frontiers in Bioengineering and Biotechnology · doi:10.3389/fbioe.2026.1832724 · Cited by 6
Progress in structure prediction and design of adaptive immune receptors
Cohen, T., Hochner, T., & Schneidman‐Duhovny, D. (2026) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2026.103326 · Cited by 4
Resource Democratization: Is Compute the Binding Constraint on AI Research?
Gelles, R., Kinoshita, V., Musser, M., & Dunham, J. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i18.29959 · Cited by 5
Bochtler, M. (2024) · BioEssays · doi:10.1002/bies.202400155 · Cited by 6
Redefining druggable targets with artificial intelligence
Akinsanya, K., AlQuraishi, M., Boija, A., Chodera, J. D., Cichońska, A., Ghassemi, M., Head, M. S., Jin, W., Kibbe, W. A., Krogan, N. J., et al. (2025) · Nature Biotechnology · doi:10.1038/s41587-025-02770-1 · Cited by 5
Self-citationKim, M., Bae, H., Jo, G., Kim, K.-W., Lee, S. J., Yoo, J., & Joo, K. (2025) · Bioinformatics · doi:10.1093/bioinformatics/btaf579 · Cited by 4
“Tools such as ColabFold (Mirdita et al. 2022) and OpenFold (Ahdritz et al. 2024) have further improved usability and computational efficiency, leading to the widespread adoption of AI-driven structure prediction.”
AlphaCRV: a pipeline for identifying accurate binder topologies in mass-modeling with AlphaFold
Guzmán-Vega, F. J., & Arold, S. T. (2024) · Bioinformatics Advances · doi:10.1093/bioadv/vbae131 · Cited by 4
“…in algorithmic efficiency have further accelerated modeling speed, enabling proteome–scale interaction predictions as a computational alternative to high-throughput experimental investigations (Mirdita et al. 2022, Wallner 2023, Yu et al. 2023, Abramson et al. 2024, Ahdritz et al. 2024).”
BackgroundZhang, I., Maysinger, D., Beus, M., Mravak, A., Yu, Z., Bakulić, M. P., Dion, P. A., Rouleau, G. A., Bonačić‐Koutecký, V., Antoine, R., et al. (2024) · Nanoscale · doi:10.1039/d4nr03512g · Cited by 4
Molecular dynamics and machine learning unlock possibilities in beauty design—A perspective
Xu, Y., Ni, H., Zhao, F., Gao, Q., Zhao, Z., Chang, C.-H., Huo, Y., Hu, S., Zhang, Y., Grovu, R. C., et al. (2025) · AIP Advances · doi:10.1063/5.0245365 · Cited by 6
xTrimoDock: Rigid Protein Docking via Cross-Modal Representation Learning and Spectral AlgorithmPreprint
Luo, Y., Li, S., Sun, Y., Wang, R., Tang, T., Hongdu, B., Cheng, X., Shi, C., Li, H., & Song, L. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.02.06.527251 · Cited by 4
“, 2021), Openfold (Ahdritz et al., 2022) and RoseTTAFold (Baek et al.”
BackgroundMethodProteinFlow: a Python Library to Pre-Process Protein Structure Data for Deep Learning ApplicationsPreprint
Kozlova, E., Valentin, A., Khadhraoui, A., & Gutierrez, D. N.-Z. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.09.25.559346 · Cited by 4
“A multitude of new model designs and architectures have been proposed to improve the learning of protein representations, including attention strategies that are tailored to the invariances and symmetries of proteins [21], graph-based representations [13, 19], andmodel recycling strategies [2, 21].”
BackgroundWu, T.-Q., Stein, R. A., Kao, T.-Y., Brown, B., & Mchaourab, H. S. (2024) · bioRxiv · doi:10.1101/2024.10.30.621127 · Cited by 3
“OpenFold is a trainable PyTorch reproduction of AlphaFold2, and its developers have verified that the implementation yields result comparable to those of the original AlphaFold2 45 .”
MethodResultInfluentialOpen-source protein structure AI aims to match AlphaFold
Naddaf, M. (2025) · Nature · doi:10.1038/d41586-025-03546-y · Cited by 4
Can We Extract Physics-like Energies from Generative Protein Diffusion Models?Preprint
Sarma, S., Truscott, H. H., Xu, D. Z., Reid, K. A., Chu, L., Chen, J. H. K., & Gray, J. J. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.11.28.690021 · Cited by 4
“The 2A1A trajectories, on the other hand, display deep and sharp oscillations and few general trends as the diffusion time approaches zero, notably not decreasing as time goes to zero.”
Kończak, J., Janusz, B., Młokosiewicz, J., Satława, T., Wróbel, S., Dudzic, P., & Krawczyk, K. (2023) · ImmunoInformatics · doi:10.1016/j.immuno.2023.100028 · Cited by 5
Wuyun, Q., Liu, Q., Ni, W., Peng, C., Zhang, Z., Zhou, X., Hu, G., Freddolino, L., & Zheng, W. H. (2025) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.70059 · Cited by 4
“In the D- I-TASSER2 pipeline (Figure 1B), the final monomer MSAs from DeepMSA2 are used as input for AlphaFold2 [2] (8-embedding), OpenFold [22], UniFold [23], ColabFold [24], RosettaFold [25], ESMFold [26], OmegaFold [27], DeepPotential [28], and AttentionPotential (an extension of DeepPotential…”
Self-citationKennedy, L. S., Harper, M., & Čuperlović‐Culf, M. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf635 · Cited by 5
“Price-171 [152] Bacterial proteins with previously misannotated EC numbers that were corrected in Price et al. [153].”
Rethinking what pLDDT really tells us about protein flexibility
Riccabona, J. R., Loeffler, J. R., Schoeder, C. T., Meiler, J., Ward, A. B., & Fernández‐Quintero, M. L. (2025) · Structure · doi:10.1016/j.str.2025.11.008 · Cited by 4
From algorithms to systems: integrating computation into drug discovery
Bradley, A. R., Rossall, A., Pairaudeau, G., & Deane, C. M. (2025) · Expert Opinion on Drug Discovery · doi:10.1080/17460441.2025.2601102 · Cited by 4
Barducci, G., Rossi, I., Codicè, F., Rollo, C., Repetto, V., Pancotti, C., Iannibelli, V., Sanavia, T., & Fariselli, P. (2026) · Communications Biology · doi:10.1038/s42003-026-09632-9 · Cited by 5
“The lack of experimental information on protein structure has led to the development of tools such as AlphaFold 28 , RoseTTaFold 29 and OpenFold 30 , which can produce high-quality protein structures but at the expense of high computational costs.”
A survey on memory-efficient transformer-based model training in AI for science
Tian, K., Qiao, L., Liu, B., Jiang, G., Li, S., & Li, D. (2026) · Frontiers of Computer Science · doi:10.1007/s11704-025-50302-6 · Cited by 13
“Based on the research of OpenFold [43], Lin et al. [46] further explored the ability of language model and proposed ESMFold.”
Black-box data: a new paradigm for biomedicine in the AI era
Naef, L., & Bronstein, M. M. (2026) · Chemical Science · doi:10.1039/d6sc01189f · Cited by 4
Mechanisms of inflammatory bowel diseases: Insights from human genetics and chemical biology
Graham, D. B., & Xavier, R. J. (2026) · Immunity · doi:10.1016/j.immuni.2026.04.002 · Cited by 3
Kaushik, R., & Re, S. (2026) · Journal of Pharmacy & Pharmaceutical Sciences · doi:10.3389/jpps.2026.16146 · Cited by 3
Assessing a computational pipeline to identify binding motifs to the α2β1 integrin
Liu, Q. Q., & Pérez, A. (2023) · Frontiers in Chemistry · doi:10.3389/fchem.2023.1107400 · Cited by 3
“OpenFold will now give users the possibility to retrain a tool equivalent to AF for new purposes.”
MethodFinke, F., Hungerland, J., Solov’yov, I. A., & Schuhmann, F. (2024) · Molecular Diversity · doi:10.1007/s11030-024-10850-8 · Cited by 3
“The second receptor was generated using AlphaFold [14, 22, 23].”
MethodKapitan, V. Y., & Choi, M. C. H. (2025) · BMC Bioinformatics · doi:10.1186/s12859-025-06185-2 · Cited by 3
“Precisely, we set: During the OpenFold [55] training process, the loss function values can vary significantly even within a single epoch.”
iQDeep: an integrated web server for protein scoring using multiscale deep learning models
Shuvo, M. H., Karim, M., & Bhattacharya, D. (2023) · Journal of Molecular Biology · doi:10.1016/j.jmb.2023.168057 · Cited by 3
Artificial Intelligence for Scientific Discovery at High-Performance Computing Scales
Lee, K. L. K., & Kumar, N. (2023) · Computer · doi:10.1109/mc.2023.3241692 · Cited by 4
RAG-ESM: Improving Pretrained Protein Language Models via Sequence Retrieval
Sgarbossa, D., & Bitbol, A.-F. (2025) · PRX Life · doi:10.1103/db1b-hy16 · Cited by 3
Benchmarking Google DeepMind’s AlphaFold 3 Performance for Protein 3D-Structure Prediction
Duma, Y., & Kyrychenko, A. V. (2024) · Kharkov University Bulletin Chemical Series · doi:10.26565/2220-637x-2024-43-01 · Cited by 4
Vinod, R., Amini, A. P., Crawford, L. A., & Yang, K. (2026) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1013925 · Cited by 2
“In contrast, protein structure models have been shown to demonstrate high generalization capabilities despite CATH-class level ablations of training data [31].”
LEP-AD: Language Embedding of Proteins and Attention to Drugs predicts drug target interactionsPreprint
Daga, A., Khan, S. A., Gómez-Cabrero, D., Hoehndorf, R., Kiani, N. A., & Tegner, J. N. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.03.14.532563 · Cited by 3
OpenProteinSet: Training data for structural biology at scale.Preprint
Ahdritz, G., Bouatta, N., Kadyan, S., Jarosch, L., Berenberg, D., Fisk, I., Watkins, A. M., Ra, S., Bonneau, R., & AlQuraishi, M. (2023) · PubMed · Cited by 3
Self-citationDrysdale, E. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.10.26.564274 · Cited by 3
“A Al-phaFold’s approach has been replicated by other researchers (12), as well as for different architectural approaches like ESMFold or RoseTTAFold (13, 14).”
MethodCerebra: a computationally efficient framework for accurate protein structure predictionPreprint
Hu, J., Wang, W., Zhang, B., Cao, Y., & Gong, H. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.02.02.578551 · Cited by 5
“For instance, OpenFold 31 , the academic implementation of AlphaFold2, requires approximately 50,000 GPU hours to train a high-accuracy model.”
Self-citationDissecting AlphaFold’s Capabilities with Limited Sequence InformationPreprint
Gut, J. A., & Lemmin, T. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.03.14.585076 · Cited by 3
“Our experimental code is openly accessible on GitHub 1 and we have contributed new methods to the OpenFold [2] project 2 .”
MethodSwapna, G. V. T., Dube, N., Roth, M. J., & Montelione, G. T. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.07.15.603529 · Cited by 3
Making use of machine learning
(2024) · Nature Chemical Biology · doi:10.1038/s41589-024-01700-7 · Cited by 3
A Structure-Based Computational Pipeline for Broad-Spectrum Antiviral DiscoveryPreprint
Castellanos, M. A., Payne, A. M., Scheen, J., MacDermott-Opeskin, H., Pulido, I., Balcomb, B. H., Griffen, E., Fearon, D., Barr, H. M., Lahav, N., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.07.29.667267 · Cited by 3
“AI-driven protein fold-76 ing models, such as AlphaFold2 ( Jumper et al., 2021 ; Evans et al., 2021 ) and RoseTTAFold ( Baek et al., 2021 ), as well as newer approaches—including AlphaFold3 Abramson et al. ( 2024 ), ESM-78 Fold ( Lin et al., 2023 ), OpenFold Ahdritz et al. ( 2022 ), and co-folding methods like Chai-1…”
Parikh, V., Foley, B., Gatlin, W., Ludwick, M., Turano, L., & Verkhivker, G. M. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.02.24.707829 · Cited by 4
AlphaFold 3 Angst: Limited Accessibility Stirs Outcry from Researchers
Lin, F. (2024) · GEN Biotechnology · doi:10.1089/genbio.2024.0029 · Cited by 2
Recent advances in the integration of protein mechanics and machine learning
Chen, Y., & Chang, S. (2024) · Extreme Mechanics Letters · doi:10.1016/j.eml.2024.102236 · Cited by 2
Phenotypic approaches for CNS drugs
Sharma, R., Oyagawa, C. R. M., Abbasi, H., Dragunow, M., & Conole, D. (2024) · Trends in Pharmacological Sciences · doi:10.1016/j.tips.2024.09.003 · Cited by 2
How to Manufacture Photonic Metamaterials
Paul, A., & Timp, G. L. (2025) · Advanced Materials Technologies · doi:10.1002/admt.202500313 · Cited by 2
Brubaker, J. K., Booth, K. E. C., Arakawa, A., Furrer, F., Ghosh, J., Sato, T., & Katzgraber, H. G. (2025) · Scientific Reports · doi:10.1038/s41598-025-05565-1 · Cited by 2
“Thanks to advances in structural biology and artificial intelligence (AI), three-dimensional models of various target proteins are readily available for SBDD [8–12].”
AFM-Fold: Rapid reconstruction of protein conformations from AFM images
Kawai, T., & Matsunaga, Y. (2026) · Biophysical Journal · doi:10.1016/j.bpj.2026.06.022 · Cited by 2
“Another approach is to use a 3D Gaussian mixture model combined with MC sampling to estimate 3D low-resolution protein structures with conformational flexibility.”
RAPID-Net: Accurate Pocket Identification for Binding-Site-Agnostic Docking
Balytskyi, Y., Hubenko, I., Balytska, A., & Kelly, C. V. (2025) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.5c01744 · Cited by 3
Regulating AI: Lessons From Scientific Computing
Wareham, J. D., Romasanta, A. K. S., Priego, L. P., & Osimo, D. (2025) · MIS Quarterly · doi:10.25300/misq/2025/19111 · Cited by 2
Kolossváry, I. (2026) · PNAS Nexus · doi:10.1093/pnasnexus/pgaf408 · Cited by 2
De novo engineering of protein interactions: Retrospective and current advances
Khramushin, A., Elizarova, E., & Correia, B. E. (2026) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2026.103240 · Cited by 2
An introduction to GitHub and its significance for AI-driven drug discovery
Hajal, A. A., Bustanji, L., Bryce, R. A., & Ghattas, M. A. (2026) · Expert Opinion on Drug Discovery · doi:10.1080/17460441.2026.2641511 · Cited by 2
“The landscape of AI-driven drug discovery on GitHub is rich and versatile, comprising specialized software tools (e.g. OpenFold [13] for protein structure prediction, GNINA [14] for molecular docking, and BAD Molecule Filter [15] for nui-sance molecules predictions), curated datasets, and…”
Brewer, W. H., Widener, P. M., Anantharaj, V. G., Wang, F., Beck, T., Shankar, M. A., & Oral, S. (2026) · AI Magazine · doi:10.1002/aaai.70056 · Cited by 3
Artificial intelligence driven innovation in tuberculosis drug discovery
Rath, N., & Arora, V. K. (2026) · Indian Journal of Tuberculosis · doi:10.1016/j.ijtb.2026.04.004 · Cited by 2
ConforFold recovers alternative protein conformations beyond MSA subsampling
Syrlybaeva, R., & Strauch, E. (2026) · Protein Science · doi:10.1002/pro.70564 · Cited by 2
“Recently, advanced deep learning methods, notably AlphaFold2/3 (Abramson et al., 2024; Bertoline et al., 2023; Jumper et al., 2021) (AF2), RoseTTAFold2/3 (Baek et al., 2021; Corley et al., 2025), OpenFold (Ahdritz et al., 2024), ESMFold (Lin et al., 2023) and OmegaFold (Wu et al., 2022) have revolutionized protein str…”
InfluentialApplications and limitations of AI tools in enzyme design
Teijeiro-Juiz, R., Egeler, N., Jamróg, G., Ringel, M., Brück, T., Geronimo, B. D., & Loll, B. (2026) · Protein Science · doi:10.1002/pro.70698 · Cited by 2
Pradeep, B. K., Chen, S., Dima, R., Hassan, M., Joachimiak, A., Kihara, D., Kloczkowski, A., Law, J., Liwo, A., Meller, J., et al. (2026) · Journal of Molecular Biology · doi:10.1016/j.jmb.2026.169949 · Cited by 2
Self-citationExpanding the scope of protein language modeling to protein-protein interactions with MSA Pairformer
Akiyama, Y., Zhang, Z., Tang, O., Kim, S., Mirdita, M., Steinegger, M., & Ovchinnikov, S. (2026) · Cell · doi:10.1016/j.cell.2026.06.029 · Cited by 3
Self-citationFifteen challenges for generative AI applications to cell biology
Dupire, L., Khan, A. A., Karaletsos, T., Kelley, S., Lundberg, E. K., Ma, J., Paull, E., Quake, S. R., Rabadán, R., Cassius, R., et al. (2026) · Cell · doi:10.1016/j.cell.2026.07.004 · Cited by 3
Self-citationVince, O., Gowers, G.-O. F., & McGibbon, S. (2024) · GEN Biotechnology · doi:10.1089/genbio.2024.0018 · Cited by 3
“One of the largest and most curated public protein databases, Uni-Prot, 47 forms the basis of all leading protein structure prediction models (e.g., AlphaFold2, ESMFold, 48 OpenFold, 49 and RoseTTaFold 50 ), protein-ligand models (e.g., NeuralPlexer 51 ), protein function models (e.g., CLEAN, 52 ProteInfer, 53 and Pro…”
Burakova, L. P., Ivanisenko, N. V., Rukosueva, N. V., Иванисенко, В. А., & Vysotski, E. S. (2024) · Life · doi:10.3390/life14111508 · Cited by 2
Yuan, Z., Shen, T., Xu, S. Y., Yu, L., Pang, B., Gan, B., Feng, Q., Ren, R., & Sun, S. (2024) · Acta Materia Medica · doi:10.15212/amm-2024-0047 · Cited by 2
“To minimize the potential influence of the template on the final results, we used the third model from OpenFold’s AF2 version [26], which is distinguished by its template omission.”
Ozheriedov, D. S., Ozheredov, S. P., Demchuk, O. M., Spivak, S. I., Blume, Y. B., & Karpov, P. A. (2025) · Cytology and Genetics · doi:10.3103/s0095452725040073 · Cited by 2
“At the time of application, the tool was trained using structures from the Protein Data Bank as well as AI structures previously predicted by AlphaFold2 (Jumper et al., 2021) and OpenFold (Ahdritz et al., 2024).”
Large-scale multi-omic biosequence transformers for modeling protein–nucleic acid interactions
Chen, S. F., Steele, R. J., Hocky, G. M., Lemeneh, B., Lad, S. P., & Oermann, E. K. (2026) · PLoS ONE · doi:10.1371/journal.pone.0341501 · Cited by 3
Self-citationEnd-to-end protein-ligand complex structure generation with diffusion-based generative modelsPreprint
Nakata, S., Mori, Y., & Tanaka, S. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.12.20.521309 · Cited by 2
GUIFold - A graphical user interface for local AlphaFold2Preprint
Kempf, G., & Cavadini, S. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.01.19.521406 · Cited by 2
“As part of the OpenFold project, AlphaFold2 has been reimplemented using the PyTorch framework (Ahdritz et al., 2022).”
MethodPAbFold: Linear Antibody Epitope Prediction using AlphaFold2
(2024) · doi:10.1101/2024.04.19.590298 · Cited by 2
“While there are many other structure prediction models other than AlphaFold2 (27, 28), including some specifically dedicated to predicting antibodies or antibody-like structures (29–32), we chose AlphaFold2 to directly test its ability to correctly identify and place epitopes into an antibody…”
PAbFold: Linear Antibody Epitope Prediction using AlphaFold2Preprint
DeRoo, J., Terry, J. S., Zhao, N., Stasevich, T. J., Snow, C. D., & Geiss, B. J. (2024) · eLife · doi:10.7554/elife.98369.1 · Cited by 2
FoldToken2: Learning compact, invariant and generative protein structure languagePreprint
Gao, Z., Tan, C., & Li, S. Z. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.06.11.598584 · Cited by 2
“For example, training well-known AlphaFold2 from scratch takes 128 TPUv3 cores for 11 days [24]; OpenFold takes 50000 GPU hours for training [2].”
MethodFoldToken4: Consistent & Hierarchical Fold LanguagePreprint
Gao, Z., Tan, C., & Li, S. Z. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.08.04.606514 · Cited by 2
“For example, training well-known AlphaFold2 from scratch takes 128 TPUv3 cores for 11 days [13]; OpenFold takes 50000 GPU hours for training [1].”
MethodQuickBind: A Light-Weight And Interpretable Molecular Docking ModelPreprint
Treyde, W., Kim, S. C., Bouatta, N., & AlQuraishi, M. (2024) · PubMed · doi:10.48550/arxiv.2410.16474 · Cited by 3
“Q UICK B IND was implemented using PyTorch [33], PyTorch Lightning [34], OpenFold [35], and the RDKit [16].”
MethodSelf-citationSwiftMHC: A High-Speed Attention Network for MHC-Bound Peptide Identification and 3D ModelingPreprint
Baakman, C. A. B., Crocioni, G., Geng, C., Rademaker, D. T., Frühbuß, D., Aarts, Y. J. M., & Xue, L. C. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.01.20.633893 · Cited by 2
Xie, L., Lu, X., Zhang, D., Xu, L., Yang, M., Jeong, Y.-S., Chang, S., & Xu, X. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.02.09.637298 · Cited by 2
“Following the success of AF2, many predictors have been generated such as RoseTTAFold 10 , ColabFold, OpenFold 11 ; ESMfold 12 , RGN2 13 , and OmegaFold 14 .”
InfluentialCombining AI structure prediction and integrative modelling for nanobody-antigen complexesPreprint
Sánchez-Marín, M., Giulini, M., & Bonvin, A. M. J. J. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.07.01.662355 · Cited by 2
Improving nanobody structure prediction with self-distillationPreprint
Ali, M., Greenig, M., Jaskółowski, M., Crnogaj, M., Smorodina, E., Zhao, H., Greiff, V., & Sormanni, P. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2025.12.01.691162 · Cited by 2
“We adopt the architecture and training objectives of ABodyBuilder3 [20, 28] to train nanobody structure predictors (details in Appendix B).”
ProFam: Open-Source Protein Family Language Modelling for Fitness Prediction and DesignPreprint
Wells, J., Hawkins‐Hooker, A., Livne, M., Lin, W., Miller, D. M., Dallago, C., Bordin, N., Paige, B., Rost, B., Orengo, C. A., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2025.12.19.695431 · Cited by 3
“The OpenProteinSet [26] is a widely used resource for defining protein families, notably for training structure prediction methods such as OpenFold [37].”
The chordata olfactory receptor database
Han, W., Bao, S., Liu, J., Wu, Y., Zeng, L., Zhang, T., Chen, N., Yao, K., Fan, S., Huang, A., et al. (2024) · Protein & Cell · doi:10.1093/procel/pwae050 · Cited by 1
“Furthermore, CORD features data on odorants, OR–odor-ants pairs and high-precision protein models (Ahdritz et al., 2024), snake diagrams (Isberg et al., 2016), BLAST (Camacho et al., 2009), profiles, WebLogos (Crooks et al., 2004), and sequence similarity networks, providing robust support for…”
Prediction and Analysis of Protein 3D Structures Using Protein Language Model and Streamlit
Srushti, C., Prathibhavani, P. M., Veena, A., & Venugopal, K. R. (2024) · International Conference on Information Management and Technology (ICIMTech) · doi:10.1109/icimtech63123.2024.10780906 · Cited by 1
“ColabFold [9] uses AF2/RosettaFold along with fast homology search for PSP, with SASA-Net [10] for PSP by inter residue distance estimates, and models with AF2 reimplementation such as HelixFold [11] OpenFold[12].”
Scrutinization on Docking against Individually Generated Target Pockets for each Ligand
Zhan, R., Zhang, W., & Huang, J. (2026) · Journal of Pharmaceutical Analysis · doi:10.1016/j.jpha.2026.101759 · Cited by 2
FAIR data for optical tweezers experiments
Halma, M. T. J., Kumar, S., van Eck, J., Abeln, S., Gates, A. J., & Wuite, G. J. L. (2025) · Biophysical Journal · doi:10.1016/j.bpj.2025.03.005 · Cited by 1
“For the case of protein structure prediction, performance of models converged above training set sizes of 2000 protein chains or above (144), establishing this as an approximate minimum threshold for training set size in protein structure prediction.”
DTIO: Data Stack for AI-driven Workflows
Bateman, K., Rajesh, N., Cernuda, J., Logan, L., Nicolae, B., Cappello, F., Sun, X., & Kougkas, A. (2025) · International Conference on Scalable Scientific Data Management (SSDBM) · doi:10.1145/3733723.3733736 · Cited by 1
AFToolkit: a framework for molecular modeling of proteins with AlphaFold-derived representations
Sindeeva, M., Telepov, A., Ivanisenko, N. V., Shashkova, T., Khrabrov, K., Tsypin, A., Kadurin, A., & Kardymon, O. L. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf324 · Cited by 1
“Deep learning has significantly advanced computational biology by providing protein modeling tools, such as AlphaFold2 [9], AlphaFold3 [10], and open-source implementations of these models [11–13].”
dl²: Detecting Communication Deadlocks in Deep Learning Jobs
Gao, Y., Luo, J., Lin, H., Zhang, H., Wu, M., & Yang, M. (2025) · ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) · doi:10.1145/3696630.3728529 · Cited by 3
The Future of Drug Discovery: An AI-Driven Perspective
Asim, M. N., Böhm, M., & Dengel, A. R. (2025) · Springer eBooks · doi:10.1007/978-3-031-93623-4_11 · Cited by 1
Tahzima, R., Charon, J., Díaz, A., Jonghe, K. D., Massart, S., Michon, T., & Vranken, W. (2025) · Frontiers in Virology · doi:10.3389/fviro.2025.1501616 · Cited by 1
Yang, L., Gao, Y. Q., Huang, Y., & Yang, J. (2025) · Scientia Sinica Chimica · doi:10.1360/ssc-2025-0169 · Cited by 1
AI in Modular Concepts of Natural Killer Cell Therapy
Vogt, Y., Kalweit, G., Alieva, M., Ullrich, E., Boedecker, J., & Kalweit, G. (2025) · Springer eBooks · doi:10.1007/978-3-662-68816-8_68-1 · Cited by 1
Computational cyclic peptide design machine learning & Rosetta based methods
Sarmeili, F., Siegler, H., Powers, A. C., & Hosseinzadeh, P. (2025) · Methods in enzymology on CD-ROM/Methods in enzymology · doi:10.1016/bs.mie.2025.09.016 · Cited by 1
Ma, X., Dey, S., Vaishnavey, S. R., Zelinski, C., Li, Q., & Chowdhury, R. (2025) · NAR Genomics and Bioinformatics · doi:10.1093/nargab/lqaf154 · Cited by 3
“Protein language-model encodings (an implicit summary of both sequence and structure) is gaining traction for protein function prediction, enzyme activity [ 29 ], distant homology detection [ 30 ], allosteric site prediction [ 31 ], subcellular localization [ 32 ] etc.”
Jung, J., Lee, T., Jung, A., Jung, H., Jung, A., Min, J., Jung, J., Zhang, B., An, J., & Hong, W. (2025) · Molecules · doi:10.3390/molecules30234516 · Cited by 2
“In contrast, models based on the AlphaFold2 architecture, like OpenFold [5], are trained primarily on 1D protein sequences and 2D multiple-sequence alignments (MSAs) to infer structural information.”
Self-citationIntegrative modeling with AlphaFold
Majila, K., Golatkar, O., & Viswanath, S. (2026) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2026.103239 · Cited by 1
The State of Peptide Detectability in Computational Proteomics and Guidelines for AI Applications
Schilling, V., Anžel, A., Doellinger, J., Lasch, P., Heider, D., & Hattab, G. (2026) · Computational and Structural Biotechnology Journal · doi:10.34133/csbj.0037 · Cited by 2
“State-of-the-art structure prediction frameworks such as AlphaFold [ 87 ], ESMFold [ 88 ], and OpenFold [ 89 ] now enable large-scale, high-accuracy inference of peptide and protein structures directly from sequence.”
AI-Driven De Novo Binder Design: From Structure Prediction to Closed-Loop Optimization
Li, X., Fan, Z., Yang, J., Jiang, R., Cao, F., Li, J., Xie, D., Meng, X., Sun, M., & Wang, L. (2026) · Computational and Structural Biotechnology Journal · doi:10.34133/csbj.0070 · Cited by 1
Metadata
- 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
- Published
- 2024-05-14
- Volume
- 21
- Issue
- 8
- Pages
- 1514-1524
- ISSN
- 1548-7091 (ISSN-L)
- 1548-7105
- 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
- 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
