Impact of AlphaFold on structure prediction of protein complexes: The CASP15‐CAPRI experiment
Open accessThe DOI matches its Crossref record; 27 of the 27 references it cites were found in a registry or an index.
Abstract
We present the results for CAPRI Round 54, the 5th joint CASP‐CAPRI protein assembly prediction challenge. The Round offered 37 targets, including 14 homodimers, 3 homo‐trimers, 13 heterodimers including 3 antibody–antigen complexes, and 7 large assemblies. On average ~70 CASP and CAPRI predictor groups, including more than 20 automatics servers, submitted models for each target. A total of 21 941 models submitted by these groups and by 15 CAPRI scorer groups were evaluated using the CAPRI model quality measures and the DockQ score consolidating these measures. The prediction performance was quantified by a weighted score based on the number of models of acceptable quality or higher submitted by each group among their five best models. Results show substantial progress achieved across a significant fraction of the 60+ participating groups. High‐quality models were produced for about 40% of the targets compared to 8% two years earlier. This remarkable improvement is due to the wide use of the AlphaFold2 and AlphaFold2‐Multimer software and the confidence metrics they provide. Notably, expanded sampling of candidate solutions by manipulating these deep learning inference engines, enriching multiple sequence alignments, or integration of advanced modeling tools, enabled top performing groups to exceed the performance of a standard AlphaFold2‐Multimer version used as a yard stick. This notwithstanding, performance remained poor for complexes with antibodies and nanobodies, where evolutionary relationships between the binding partners are lacking, and for complexes featuring conformational flexibility, clearly indicating that the prediction of protein complexes remains a challenging problem.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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
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
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
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
UniProt: the universal protein knowledgebase in 2021
The UniProt Consortium, Bateman, A., Martin, M.-J., Orchard, S., Magrane, M., Agivetova, R., Ahmad, S., Alpi, E., Bowler-Barnett, E. H., Britto, R., et al. (2021) · Nucleic Acids Research · doi:10.1093/nar/gkaa1100 · Cited by 6,372
Self-citationAccurate prediction of protein structures and interactions using a 3-track network
Baek, M., DiMaio, F., Anishchenko, I., Dauparas, J., Ovchinnikov, S., Lee, G. R., Wang, J., Cong, Q., Kinch, L. N., Schaeffer, R. D., et al. (2021) · doi:10.1101/2021.06.14.448402 · Cited by 5,549 · cited 1×
“More specifically, the extent to which the power of AlphaFold and other DL-based prediction methods, such as RoseTTAFold, 35 will be harnessed by the community to produce a major leap in performance over more classical modeling approaches in the context of blind predictions.”
Network medicine: a network-based approach to human disease
Barabási, A.-L., Gulbahce, N., & Loscalzo, J. (2011) · Nature Reviews Genetics · doi:10.1038/nrg2918 · Cited by 5,034 · cited 1×
“2,3 Characterizing these interactions and elucidating their functions at the molecular and cellular levels have therefore been important goals in molecular biology and medicine.”
Protein complex prediction with AlphaFold-Multimer
Evans, R., O’Neill, M., Pritzel, A., Antropova, N., Senior, A., Green, T., Žídek, A., Bates, R., Blackwell, S., Yim, J., et al. (2021) · doi:10.1101/2021.10.04.463034 · Cited by 3,808 · cited 3×
“The performance of both approaches heavily relied on the AF2/AF2-M confidence metrics (pLDDT, iPTM, pTM) 17,33 to prioritize and rank models, a practice backed by recent findings that these metrics closely estimate the true quality of the candidate structures, outperforming other state-of-the-art…”
Announcing the worldwide Protein Data Bank
Berman, H., Henrick, K., & Nakamura, H. (2003) · Nature Structural & Molecular Biology · doi:10.1038/nsb1203-980 · Cited by 3,093
The Cell as a Collection of Protein Machines: Preparing the Next Generation of Molecular Biologists
Alberts, B. (1998) · Cell · doi:10.1016/s0092-8674(00)80922-8 · Cited by 1,411
Improved prediction of protein-protein interactions using AlphaFold2
Bryant, P., Pozzati, G., & Elofsson, A. (2021) · doi:10.1101/2021.09.15.460468 · Cited by 1,022 · cited 1×
“32 Better performance, albeit still not reaching the high model accuracy obtained for single chain proteins, was achieved for AlphaFold2-Multimer (AF2-M), the more recent inference engine of AlphaFold, directly trained on protein complexes from the PDB. 33 Approaches have also been proposed to…”
Cited by
Critical assessment of methods of protein structure prediction ( CASP )—Round XV
Kryshtafovych, A., Schwede, T., Topf, M., Fidelis, K., & Moult, J. (2023) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.26617 · Cited by 148
“Agreement with experiment also tended to be lower for the few NMR targets.”
BackgroundYin, R., & Pierce, B. G. (2024) · Protein Science · doi:10.1002/pro.4865 · Cited by 145
Self-citationMassiveFold: unveiling AlphaFold’s hidden potential with optimized and parallelized massive sampling
Raouraoua, N., Mirabello, C., Véry, T., Blanchet, C., Wallner, B., Lensink, M. F., & Brysbaert, G. (2024) · Nature Computational Science · doi:10.1038/s43588-024-00714-4 · Cited by 51
Self-citationGiulini, M., Schneider, C., Cutting, D., Desai, N., Deane, C. M., & Bonvin, A. M. J. J. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btae583 · Cited by 41
Self-citationA fully automated benchmarking suite to compare macromolecular complexes
Studer, G., Robin, X., Bienert, S., Durairaj, J., Škrinjar, P., Tauriello, G., Waterhouse, A. M., & Schwede, T. (2026) · Nature Methods · doi:10.1038/s41592-025-02973-z · Cited by 10
“Historically, CASP uses GDT_TS (total score), which averages GDT scores with distance thresholds 1, 2, 4 and 8 Å .”
InfluentialSeaMoon: From protein language models to continuous structural heterogeneity
Lombard, V., Timsit, D., Grudinin, S., & Laine, E. (2025) · Structure · doi:10.1016/j.str.2025.06.010 · Cited by 8
Towards a Greener AlphaFold2 Protocol for Antibody–Antigen Modeling: Insights From CAPRI Round 55
Savaş, B., Yılmazbilek, İ., Özsan, A., & Karaca, E. (2026) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.26820 · Cited by 4
Thomas, D. P. G., Garcia Fernandez, C. M., Haydarlou, R., & Feenstra, K. A. (2025) · Scientific Reports · doi:10.1038/s41598-025-88445-y · Cited by 4
Olechnovič, K., & Grudinin, S. (2025) · Journal of Computational Chemistry · doi:10.1002/jcc.70178 · Cited by 3
Xia, Y., Pu, Y., Wang, S., Zhuang, J., Liu, D., Hou, M., & Zhang, G. (2025) · Journal of Molecular Biology · doi:10.1016/j.jmb.2025.169128 · Cited by 2
Xia, Y., Pu, Y., Wang, S., Zhuang, J., Liu, D., Hou, M., & Zhang, G. (2024) · doi:10.1101/2024.11.22.624951 · Cited by 2
Metadata
- Authors
- Lensink, M. F., Brysbaert, G., Raouraoua, N., Bates, P. A., Giulini, M., Honorato, R. V., van Noort, C., Teixeira, J. M. C., Bonvin, A. M. J. J., Kong, R., Shi, H., Lu, X., Chang, S., Liu, J., Guo, Z., Chen, X., Morehead, A., Roy, R. S., Wu, T., Giri, N., Quadir, F., Chen, C., Cheng, J., Del Carpio, C. A., Ichiishi, E., Rodriguez‐Lumbreras, L. A., Fernandez‐Recio, J., Harmalkar, A., Chu, L., Canner, S., Smanta, R., Gray, J. J., Li, H., Lin, P., He, J., Tao, H., Huang, S., Roel‐Touris, J., Jimenez‐Garcia, B., Christoffer, C. W., Jain, A. J., Kagaya, Y., Kannan, H., Nakamura, T., Terashi, G., Verburgt, J. C., Zhang, Y., Zhang, Z., Fujuta, H., Sekijima, M., Kihara, D., Khan, O., Kotelnikov, S., Ghani, U., Padhorny, D., Beglov, D., Vajda, S., Kozakov, D., Negi, S. S., Ricciardelli, T., Barradas‐Bautista, D., Cao, Z., Chawla, M., Cavallo, L., Oliva, R., Yin, R., Cheung, M., Guest, J. D., Lee, J., Pierce, B. G., Shor, B., Cohen, T., Halfon, M., Schneidman‐Duhovny, D., Zhu, S., Yin, R., Sun, Y., Shen, Y., Maszota‐Zieleniak, M., Bojarski, K. K., Lubecka, E. A., Marcisz, M., Danielsson, A., Dziadek, L., Gaardlos, M., Gieldon, A., Liwo, A., Samsonov, S. A., Slusarz, R., Zieba, K., Sieradzan, A. K., Czaplewski, C., Kobayashi, S., Miyakawa, Y., Kiyota, Y., Takeda‐Shitaka, M., Olechnovic, K., Valancauskas, L., Dapkunas, J., Venclovas, C., Wallner, B., Yang, L., Hou, C., He, X., Guo, S., Jiang, S., Ma, X., Duan, R., Qui, L., Xu, X., Zou, X., Velankar, S., & Wodak, S. J.
- Venue
- Proteins: Structure, Function, and Bioinformatics
- Publisher
- Wiley
- Published
- 2023-10-31
- Volume
- 91
- Issue
- 12
- Pages
- 1658-1683
- ISSN
- 0887-3585 (ISSN-L)
- 1097-0134
- OpenAlex
- W4388034457
- Cited by
- 77
- References
- 27
- FWCI
- 5.75
- Type
- journal article
- Language
- en
- Open access
- hybrid
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.