Description of the experiment
CASP (Critical Assessment of Structure Prediction) is a community wide experiment to determine and advance the state of the art in computational structural biology. Every two years, participants are invited to submit models for a set of macromolecules and their complexes (proteins, RNA, ligands) for which the experimental structures are not yet public. In the latest CASP round, CASP16 in 2024, over 100 research groups from around the world submitted more than 120,000 models on 100+ targets in nine broad prediction categories. Independent assessors then compare the models with experiment. Assessments and results are posted at this website and published in a special issue of a scientific journal (lately PROTEINS - (check the latest CASP16 issue here).
The goals of CASP are to provide rigorous assessment of computational methods for modeling macromolecular structures and complexes so as to drive advances in the field. Recent CASPs saw enormous jumps in the accuracy of computed structures, first in CASP14 (2020) for single proteins and domains, with many models competitive in accuracy with experiment, and second, in CASP15 (2022), with a large increase in the accuracy of protein complexes. These advances are primarily the result of the successful application of deep learning methods, particularly AlphaFold2 and other methods built around it. But results from CASP16 (2024) suggest a performance plateau.
The primary goal for the 2026 CASP17 experiment is to catalyze breakthroughs in areas where deep learning has yet to deliver and particularly where success has major practical implications. CASP17 modeling categories are chosen accordingly.
-
Immune Complexes:
This is an area of major practical importance, and also one that in CASP is a major failure area for current deep learning methods. However, two promising approaches have been seen in recent CASPs. To nurture progress, we are introducing immune complexes as a specific category in CASP17, aiming to provide a rich and varied set of non-homologous targets spanning antibody-antigen, nanobody-antigen complexes, and T-cell receptor complexes.
-
Organic Ligand-Protein Complexes:
These complexes have obvious importance for the development of new small-molecule drugs. The most recent CASP16 (2024) showed that current deep learning results often fall short of the experimental accuracy, especially where there is little homologous or analogous data available. Thus, we will again include this as a specific target area in CASP17, aiming to obtain targets representing a wide range of realistic conditions.
-
Nucleic Acids and Complexes:
Despite claims that deep learning methods have solved the problem of computing nucleic acid structures, CASP16 and a subsequent Kaggle challenge showed that these methods are usually no better than classical approaches and that both fail badly in the absence of homologous structural information. But new deep learning methods are now appearing at a fast pace. To assess these, CASP17 will include a diversity of non-homologous RNA/DNA structures and protein-nucleic acid complexes. As before, this category will be in collaboration with RNA Puzzles, and will be co-ordinated with RNA Kaggle challenges. Because of the speed with which some experimental structures may be obtained we plan to shorten the prediction window for this kind of target.
-
Conformational Ensembles:
Testing methods for computing ensembles of structures is a major expansion area for CASP. CASP17 will again include two main types of ensemble target: First, those where there are two or a few discrete conformations, each with a well-determined experimental structure, that can be assessed in the conventional way. Second, targets where there are multiple lower resolution experimental datasets available, such as cryo-tomography; SAXS; NMR (RDC, chemical shifts and other data); FRET; and cross-linking. For these, assessment will be based on agreement between the low-resolution experimental data and corresponding values calculated from submitted structural ensembles.
-
Difficult Protein Structures and Complexes:
Recent CASPs have shown that in many cases, current deep learning methods deliver high-accuracy structures for single proteins and complexes. But there are critical weaknesses. To help address these, CASP17 will focus on performance in the following areas:
- Membrane
proteins.
- Proteins
and complexes with weak evolutionary information such as those with viral
or parasite origin, "shallow" sequence alignments and recently
evolved interfaces.
- Large
proteins and complexes with complicated stoichiometry or arrangement of subunits
(>1,000 amino acids).
For protein complexes we will again work in close collaboration with our CAPRI partners.
Note: CASP17 protein targets will be initially released without stoichiometry information. All multimeric targets will be re-released in a second modeling stage, with the experimental stoichiometry data provided. For this second stage, the MassiveFold team will also provide large sets of AF-based models after the server deadline (typically 3 days after the second-stage target release).
-
Accuracy Estimation:
A structural model that is not accompanied by reliable and detailed accuracy methods is of very limited value. Estimates for protein structures were a success story in CASP even before the introduction of deep learning methods. Now CASP results have shown that accuracy estimates provided by model builders are consistently reliable. For third-party methods (those that estimate accuracy for models produced by others) we will no longer include single proteins, but will include protein complex interfaces from the MassiveFold and CASP stage-2 models. New this CASP is self-assessment of accuracy for nucleic acid structures and complexes, and for protein-ligand complexes.
- March 30, 2026 - Start of the registration for CASP17 prediction experiment.
- April 13 - Start of the testing of server connectivity ("dry run" for server predictors). New in CASP17 - we will be accepting submissions only through the web form or server API. No email submissions.
- April 27 - May 1 - First week of releasing CASP17 targets.
- June/July - Early bird registration for the December CASP17 conference.
- July 31 - Last date for releasing targets.
- August 31 - End of the modeling season.
- Early September - Collection of CASP17 methods abstracts. Note: no abstract - no recognition by the assessors and no talk at the conference.
- August-October - Evaluation of predictions.
- November - Invitations to groups with the most accurate models
and the most interesting methods to give talks at the CASP17 conference.
- Second part of November - Program of the conference finalized.
- December 2026 - CASP17 Conference. Details to follow.
Participation is open to all.
If you are new to CASP and don't have an account with the Prediction Center, you will have to
register with the Prediction Center first and only then proceed to
CASP17 registration page, which will be available here on March 30, 2026.
If you already have an account with the Prediction Center, you can go directly to the
CASP17 registration page.
Please check, though, that your basic registration information is
current. If it has changed - please update it through the My Personal
Data link from the main Menu.
Participants with servers are requested to register before April 14 as
we are planning to start checking servers' format and connectivity thereafter.
CASP17 modeling targets will be announced through the Target List page from
the main CASP17 webpage.
The success of CASP is dependent on the generous help of the experimental community in providing targets. As in previous CASPs, protein crystallographers, NMR spectroscopists and cryo-EM scientists are asked to provide details of structures they expect to have made public before September 2026.
For the CASP17 experiment we are seeking targets with specific features mentioned in the Categories section. In particular, we need immune complexes, single proteins that are viral, membrane-related, large or those with shallow MSA; large assemblies; RNA/DNA and complexes of thereof with proteins. For the protein-organic ligand category, we are targeting sets of drug design-related complexes, but other complexes are also welcome. Success of the macromolecular ensembles category depends on the availability of targets with multiple experimentally observed conformations or multiple low resolution data types. The last day for suggesting CASP targets is July 10, 2026.
A target submission form is available here.
Details on the target collection and release procedures are available at our
Q&A page.
Models can be submitted through the Prediction Submission form available from this website
or server API. No submissions through email will be accepted. Please comply with the
submission procedures and format.
As is the practice in CASP, assessment of the results will be made by the independent assessor teams. Assessment criteria will be based on those previously developed in CASP, but assessors may add new metrics they consider appropriate. Where possible, results will also be evaluated using criteria from the previous CASP, so the effects of any changes in criteria can be appreciated.
The CASP17 Assessors are as follows:
- Immune complexes - TBA
- Difficult monomers and complexes - Ronan Keegan (STFC Rutherford Appleton Laboratory, UK)
- Ligands - Janani (Jay) Durairaj (University of Basel, Switzerland)
- RNA - Marta Szachniuk (Poznan University of Technology, Poland) and Eric Westhof (Universite de Strasbourg, France)
- Ensembles - TBA
Click here
for the list of assessors in all CASPs held so far.
In accordance with CASP policy, assessors cannot take part in the relevant parts of the experiment as predictors. Participants must not contact assessors directly with queries, but rather these should be sent to the
email
address.
All CASP predictions and results of numerical evaluation will be made available through
this web site shortly before the meeting.
The proceedings will be published in a scientific journal
(see
publications of previous experiments).
All participants are required to describe their methods in the abstracts,
which will be published at our web site in late October.
No talks at the conference will be offered to modelers without submitted abstracts.
The conference to discuss results of the CASP17 experiment will be held
in December 2026. Details to follow.
Registration for the meeting will open in early June.
John Moult, CASP chair and founder; IBBR, University of Maryland, USA
Krzysztof Fidelis, founder, University of California, Davis, USA
Andriy Kryshtafovych, University of California, Davis, USA
Torsten Schwede, University of Basel, Switzerland
Maya Topf, Centre for Structural Systems Biology, Hamburg, Germany
Minkyung Baek, Seoul University, South Korea
David Baker, University of Washington, USA
Charlotte Dean, University of Oxford, UK
Nick Grishin, University of Texas, USA
Andrzej Joachimiak, Argonne National Lab, USA
David Jones, University College, London, UK
John Jumper, Google Deepmind, London, UK
|
