CoChat
Sign Up
PreprintCited by 2Protein Structure and DynamicsEnzyme Structure and FunctionMachine Learning in Bioinformatics

The DOI matches its Crossref record; 8 of the 8 references it cites were found in a registry or an index.

Assessing AF2’s ability to predict structural ensembles of proteins

Open access
Riccabona, Jakob R., Spoendlin, Fabian C., Fischer, Anna-Lena M., Loeffler, Johannes R., Quoika, Patrick K., Jenkins, Timothy P., Ferguson, James A., Smorodina, Eva, Laustsen, Andreas H., Greiff, Victor, Forli, Stefano, Ward, Andrew B., Deane, Charlotte M., Fernández-Quintero, Monica L.·2024·bioRxiv (Cold Spring Harbor Laboratory)
DOI 10.1101/2024.04.16.589792

Preprint — published as 10.1016/j.str.2024.09.001 · Paper page

Open this paper in CoChat — free accountOpen PDF

Abstract

Recent breakthroughs in protein structure prediction have enhanced the precision and speed at which protein configurations can be determined, setting new benchmarks for accuracy and efficiency in the field. However, the fundamental mechanisms of biological processes at a molecular level are often connected to conformational changes of proteins. Molecular dynamics (MD) simulations serve as a crucial tool for capturing the conformational space of proteins, providing valuable insights into their structural fluctuations. However, the scope of MD simulations is often limited by the accessible timescales and the computational resources available, posing challenges to comprehensively exploring protein behaviors. Recently emerging approaches have focused on expanding the capability of AlphaFold2 (AF2) to predict conformational substates of protein structures by manipulating the input multiple sequence alignment (MSA). These approaches operate under the assumption that the MSA also contains information about the heterogeneity of protein structures. Here, we benchmark the performance of various workflows that have adapted AF2 for ensemble prediction focusing on the subsampling of the MSA as implemented in ColabFold and compare the obtained structures with ensembles obtained from MD simulations and NMR. As test cases, we chose four proteins namely the bovine pancreatic inhibitor protein (BPTI), thrombin and two antigen binding fragments (antibody Fv and nanobody), for which reliable experimentally validated structural information (X-ray and/or NMR) was available. Thus, we provide an overview of the levels of performance and accessible timescales that can currently be achieved with machine learning (ML) based ensemble generation. In three out of the four test cases, we find structural variations fall within the predicted ensembles. Nevertheless, significant minima of the free energy surfaces remain undetected. This study highlights the possibilities and pitfalls when generating ensembles with AF2 and thus may guide the development of future tools while informing upon the results of currently available applications.

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.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1101/2024.04.16.589792

Cites

8 of 8 references verified0 citation errors · 0 not checkable

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.

A few free checks a day · papers already verified open instantly · results are public

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

Flagged first, then by citations

Research works printed but not linked: 80

  1. 10.1103/PhysRevLett.129.238101 · DOI · Paper page
  2. 10.1126/science.146.3651.1593 · DOI
  3. 10.1038/nature06522 · DOI · Paper page
  4. 10.1063/1.1755656 · DOI
  5. Chakravarty, D. & Porter, L. L . AlphaFold2 fails to predict protein fold switching. Protein Science 31, (2022). · DOI · Paper page
  6. Chakravarty, D. , Schafer, J. W. & Porter, L. L . Distinguishing features of fold-switching proteins. Protein Science 32, (2023). · DOI
  7. 10.1038/s41594-021-00714-2 · DOI · Paper page
  8. Monteiro da Silva, G. , Cui, J. Y. , Dalgarno, D. C. , Lisi, G. P. & Rubenstein, B. M. Predicting Relative Populations of Protein Conformations without a Physics Engine Using AlphaFold 2. Preprint (2023). · DOI
  9. 10.1038/s41467-023-41237-2 · DOI
  10. 10.1016/S0022-2836(84)80006-6 · DOI
  11. 10.1107/S0907444998008725 · DOI
  12. 10.1107/S0907444998008725 · DOI
  13. 10.1016/0014-5793(75)80504-7 · DOI
  14. 10.1126/science.1948083 · DOI
  15. 10.1021/ja0775873 · DOI
  16. Zhang, W. , Nielsen, C. B. & Hansen, P. E . Solution structure of a truncated bovine pancreatic trypsin inhibitor, 3–58 BPTI. (2004). · DOI
  17. 10.1126/science.1187409 · DOI · Paper page
  18. 2020, Protein-Protein Binding as a Two-Step Mechanism: Preselection of Encounter Poses during the Binding of BPTI and Trypsin, Biophys J, 119, 652, 10.1016/j.bpj.2020.06.032 · DOI
  19. 10.1021/ja0367389 · DOI
  20. 10.1021/bi00107a001 · DOI
  21. 2006, An Overview of the Structure and Function of Thrombin, Semin Thromb Hemost, 32, 003, 10.1055/s-2006-939550 · DOI
  22. 10.1016/j.mam.2008.01.001 · DOI
  23. 1989, The refined 1.9 A crystal structure of human alpha-thrombin: interaction with D-Phe-Pro-Arg chloromethylketone and significance of the Tyr-Pro-Pro-Trp insertion segment, EMBO J, 8, 3467, 10.1002/j.1460-2075.1989.tb08511.x · DOI
  24. 2006, The Structure of Thrombin: A Janus-Headed Proteinase, Semin Thromb Hemost, 32, 016, 10.1055/s-2006-939551 · DOI
  25. 10.1038/s41598-020-57822-0 · DOI
  26. 10.1074/jbc.M608600200 · DOI
  27. Pain, C. et al. Structure of a single domain camelid antibody fragment cAb-H7S in complex with the BlaP beta-lactamase from Bacillus licheniformis. (2014). · DOI
  28. 2019, CDR-H3 loop ensemble in solution – conformational selection upon antibody binding, MAbs, 11, 1077, 10.1080/19420862.2019.1618676 · DOI
  29. 10.1126/science.1546293 · DOI
  30. 10.1038/S41592-022-01488-1 · DOI · Paper page
  31. 10.1038/nbt.3988 · DOI · Paper page
  32. 10.1093/nar/gkab1050 · DOI
  33. 10.1093/nar/gkt1043 · DOI · Paper page
  34. Spoendlin, F. C. et al. Improved computational epitope profiling using structural models identifies a broader diversity of antibodies that bind to the same epitope. Front Mol Biosci 10, (2023). · DOI · Paper page
  35. 2022, Nanobody Paratope Ensembles in Solution Characterized by MD Simulations and NMR, Int J Mol Sci, 23, 5419, 10.3390/ijms23105419 · DOI
  36. 2015, GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers, SoftwareX, 1–2, 19
  37. 10.1093/bioinformatics/btt055 · DOI
  38. 10.1016/j.cpc.2013.09.018 · DOI
  39. 2008, Metadynamics: a method to simulate rare events and reconstruct the free energy in biophysics, chemistry and material science, Reports on Progress in Physics, 71, 126601, 10.1088/0034-4885/71/12/126601 · DOI · Paper page
  40. 10.1103/PhysRevLett.100.020603 · DOI · Paper page
  41. 2015, Bias-Exchange Metadynamics Simulations: An Efficient Strategy for the Analysis of Conduction and Selectivity in Ion Channels, J Chem Theory Comput, 11, 1896, 10.1021/ct501053x · DOI
  42. Fernández-Quintero, M. L. et al. Characterizing the Diversity of the CDR-H3 Loop Conformational Ensembles in Relationship to Antibody Binding Properties. Front Immunol 9, (2019). · DOI
  43. Fernández-Quintero, M. L. , Pomarici, N. D. , Loeffler, J. R. , Seidler, C. A. & Liedl, K. R . T-Cell Receptor CDR3 Loop Conformations in Solution Shift the Relative Vα-Vβ Domain Distributions. Front Immunol 11, (2020). · DOI
  44. 10.1021/ct700119m · DOI
  45. 10.1021/ct400341p · DOI
  46. 10.1021/ct400314y · DOI
  47. 10.1002/1096-987X(20010415)22:5<501::AID-JCC1021>3.0.CO;2-V · DOI
  48. 10.1063/1.448118 · DOI
  49. 1975, Generalized Langevin equation approach for atom/solid-surface scattering: Inelastic studies, J Chem Phys, 63, 4908, 10.1063/1.431234 · DOI
  50. 10.1021/acs.jctc.5b00743 · DOI
  51. 10.1109/TIT.1982.1056489 · DOI · Paper page
  52. 10.1007/s11634-013-0134-6 · DOI
  53. 10.1016/0003-9861(80)90406-3 · DOI
  54. 10.1021/bi00162a008 · DOI
  55. Mikolajek, H. et al. Correlation between the binding affinity and the conformational entropy of nanobody SARS-CoV-2 spike protein complexes. Proceedings of the National Academy of Sciences 119, (2022). · DOI
  56. 2023, Orally delivered single-domain antibodies against gastrointestinal pathogens, Trends Biotechnol, 41, 875, 10.1016/j.tibtech.2023.01.015 · DOI
  57. 2024, Structural trends in antibody-antigen binding interfaces: a computational analysis of 1833 experimentally determined 3D structures, Comput Struct Biotechnol J, 23, 199, 10.1016/j.csbj.2023.11.056 · DOI
  58. Mejias-Gomez, O. et al. A window into the human immune system: comprehensive characterization of the complexity of antibody complementary-determining regions in functional antibodies. MAbs 15, (2023). · DOI
  59. Gordon, G. L. et al. A comparison of the binding sites of antibodies and single-domain antibodies. Front Immunol 14, (2023). · DOI
  60. Fernández-Quintero, M. L. et al. Challenges in antibody structure prediction. MAbs 15, (2023). · DOI · Paper page
  61. 10.1016/j.str.2014.11.010 · DOI
  62. 10.1038/s41467-023-36699-3 · DOI
  63. 10.1016/j.sbi.2023.102645 · DOI · Paper page
  64. 10.1146/annurev.bi.52.070183.001403 · DOI
  65. 10.1038/nature06522 · DOI · Paper page
  66. 2022, Conformational Entropy as a Potential Liability of Computationally Designed Antibodies, Biomolecules, 12, 718, 10.3390/biom12050718 · DOI
  67. 10.1038/347631a0 · DOI
  68. 2023, AlphaFold2-RAVE: From Sequence to Boltzmann Ranking, J Chem Theory Comput, 19, 4351, 10.1021/acs.jctc.3c00290 · DOI
  69. Ray, D. , Trizio, E. & Parrinello, M . Deep learning collective variables from transition path ensemble. J Chem Phys 158, (2023). · DOI
  70. 2022, Accelerating Protein Folding Molecular Dynamics Using Inter-Residue Distances from Machine Learning Servers, J Chem Theory Comput, 18, 1929, 10.1021/acs.jctc.1c00916 · DOI
  71. 10.1039/D2RA03660F · DOI
  72. 2022, Multitask Machine Learning of Collective Variables for Enhanced Sampling of Rare Events, J Chem Theory Comput, 18, 2341, 10.1021/acs.jctc.1c00143 · DOI
  73. Bonati, L. , Piccini, G. & Parrinello, M . Deep learning the slow modes for rare events sampling. Proceedings of the National Academy of Sciences 118, (2021). · DOI
  74. 10.1021/acs.jctc.3c01081 · DOI · Paper page
  75. Gu, X. , Aranganathan, A. & Tiwary, P. Empowering AlphaFold2 for protein conformation selective drug discovery with AlphaFold2-RAVE. (2024). · DOI
  76. 10.1103/PhysRevLett.129.238101 · DOI · Paper page
  77. 2021, The case for post-predictional modifications in the AlphaFold Protein Structure Database, Nat Struct Mol Biol, 28, 869, 10.1038/s41594-021-00680-9 · DOI
  78. 2023, ImmuneBuilder: Deep-Learning models for predicting the structures of immune proteins, Commun Biol, 6, 575, 10.1038/s42003-023-04927-7 · DOI
  79. Jiarui Lu, B. Z. J. T. Score-based Enhanced Sampling for Protein Molecular Dynamics. ICML 2023 Workshop on Structured Probabilistic Inference & Generative Modeling Preprint at (2023).
  80. 10.1101/2023.08.01.551540 · DOI

Metadata

Authors
Riccabona, J. R., Spoendlin, F. C., Fischer, A.-L. M., Loeffler, J. R., Quoika, P. K., Jenkins, T. P., Ferguson, J. A., Smorodina, E., Laustsen, A. H., Greiff, V., Forli, S., Ward, A. B., Deane, C. M., & Fernández-Quintero, M. L.
Affiliations
  • Universität Innsbruck (AT)
  • University of Oxford (GB)
  • Technical University of Munich (DE)
  • Technical University of Denmark (DK)
  • Scripps Research Institute (US)
  • University of Oslo (NO)
ORCID
Venue
bioRxiv (Cold Spring Harbor Laboratory)
Publisher
openRxiv
Published
2024-04-17
DOI
10.1101/2024.04.16.589792
OpenAlex
W4394928557
Cited by
2
References
8
Citations by year
2025: 1 · 2024: 1
Type
posted content
Language
en
Open access
Licence
cc-by
Locations
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
  • Technical University of Denmark, DTU Orbit (Technical University of Denmark, DTU) (submittedVersion, cc-by-nc-nd)
Funders
  • European High Performance Computing Joint Undertaking
  • Bill and Melinda Gates Foundation (INV-004923)
  • Leona M. and Harry B. Helmsley Charitable Trust (#2019PG-T1D011)
  • Villum Fonden (00025302)
  • European Commission (HORIZON2020, 825821, 850974, 101125630)
  • Norges Forskningsråd (331890, 311341, 300 740)
  • Engineering and Physical Sciences Research Council (EP/S024093, EP/S024093/1)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Structural Biology › Enzyme Structure and Function
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
Keywords
AlphaFold2, protein structural ensembles, conformational heterogeneity, ColabFold, molecular dynamics simulation, protein conformational change, aprotinin, thrombin, nanobodies

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