CoChat
Sign Up
Cited by 140Protein Structure and DynamicsEnzyme Structure and FunctionMachine Learning in Bioinformatics

Improved AlphaFold modeling with implicit experimental information

Open access
Terwilliger, Thomas C., Poon, Billy K., Afonine, Pavel V., Schlicksup, Christopher J., Croll, Tristan I., Millán, Claudia, Richardson, Jane. S., Read, Randy J. +1 more·2022·Nature Methods
DOI 10.1038/s41592-022-01645-6PMC PMC9636017

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

Machine-learning prediction algorithms such as AlphaFold and RoseTTAFold can create remarkably accurate protein models, but these models usually have some regions that are predicted with low confidence or poor accuracy. We hypothesized that by implicitly including new experimental information such as a density map, a greater portion of a model could be predicted accurately, and that this might synergistically improve parts of the model that were not fully addressed by either machine learning or experiment alone. An iterative procedure was developed in which AlphaFold models are automatically rebuilt on the basis of experimental density maps and the rebuilt models are used as templates in new AlphaFold predictions. We show that including experimental information improves prediction beyond the improvement obtained with simple rebuilding guided by the experimental data. This procedure for AlphaFold modeling with density has been incorporated into an automated procedure for interpretation of crystallographic and electron cryo-microscopy maps.

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 · 10.1038/s41592-022-01645-6

Cites

6 of 6 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.

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

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?

  • An iterative AF2-integrated molecular replacement procedure has been developed, using AF models in the initial structure-solution cycle followed by data-guided cycles of AF prediction and model rebuilding.

    “Recently, an AF2 integrated iterative procedure for molecular replacement has been developed where AF models are utilized during the initial structure-solution cycle, followed by data guided cycles of AF structure prediction and model rebuilding60,69.”

    From The power and pitfalls of AlphaFold2 for structure prediction beyond rigid globular proteins (2024) · cites this paper as [60]

  • AF models have been used extensively with single-particle cryo-EM data.

    “AF models have also been used extensively together with single particle cryo-EM data60,76,95,96”

    From The power and pitfalls of AlphaFold2 for structure prediction beyond rigid globular proteins (2024) · cites this paper as [60]

  • Iterative rebuilding of AF2 models based on cryo-EM data enables faster and more robust structure modeling than ab initio model building.

    “Like for X-ray diffraction studies, implicit incorporation of AF2 models, which are iteratively rebuilt on the basis of cryo-EM data, enables swift and robust structure modeling relative to ab initio model building60,98.”

    From The power and pitfalls of AlphaFold2 for structure prediction beyond rigid globular proteins (2024) · cites this paper as [60]

Self-check

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

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

Cites

Flagged first, then by citations

Research works printed but not linked: 27

  1. [2]Baek M Accurate prediction of protein structures and interactions using a three-track neural network Science 2021 373 871 876 10.1126/science.abj8754 34282049 PMC7612213 · DOI · Paper page
  2. [3]David A Islam S Tankhilevich E Sternberg MJE The AlphaFold database of protein structures: a biologist’s guide J. Mol. Biol. 2022 434 167336 10.1016/j.jmb.2021.167336 34757056 PMC8783046 · DOI · Paper page
  3. [4]Perrakis A Sixma TK AI revolutions in biology EMBO Rep. 2021 22 e54046 10.15252/embr.202154046 34668287 PMC8567224 · DOI · Paper page
  4. [6]Cramer P AlphaFold2 and the future of structural biology Nat. Struct. Mol. Biol. 2021 28 704 705 10.1038/s41594-021-00650-1 34376855 · DOI
  5. [8]Gupta, M. et al. CryoEM and AI reveal a structure of SARS-CoV-2 Nsp2, a multifunctional protein involved in key host processes. Preprint at bioRxiv10.1101/2021.05.10.443524 (2021).
  6. [9]AlQuraishi M Machine learning in protein structure prediction Curr. Opin. Chem. Biol. 2021 65 1 8 10.1016/j.cbpa.2021.04.005 34015749 · DOI
  7. [11]Dunker AK Intrinsically disordered protein J. Mol. Graph. Modell. 2001 19 26 59 10.1016/S1093-3263(00)00138-8 11381529 · DOI · Paper page
  8. [13]Mariani V Biasini M Barbato A Schwede T lDDT: a local superposition-free score for comparing protein structures and models using distance difference tests Bioinformatics 2013 29 2722 2728 10.1093/bioinformatics/btt473 23986568 PMC3799472 · DOI · Paper page
  9. [14]Ju F CopulaNet: learning residue co-evolution directly from multiple sequence alignment for protein structure prediction Nat. Commun. 2021 12 2535 10.1038/s41467-021-22869-8 33953201 PMC8100175 · DOI
  10. [15]Perrakis A Morris R Lamzin VS Automated protein model building combined with iterative structure refinement Nat. Struct. Biol. 1999 6 458 463 10.1038/8263 10331874 · DOI
  11. [16]Wang L Ultrapotent antibodies against diverse and highly transmissible SARS-CoV-2 variants Science 2021 373 eabh1766 10.1126/science.abh1766 34210892 PMC9269068 · DOI
  12. [17]Cater RJ Structural basis of omega-3 fatty acid transport across the blood–brain barrier Nature 2021 595 315 319 10.1038/s41586-021-03650-9 34135507 PMC8266758 · DOI
  13. [18]Terwilliger TC Adams PD Afonine PV Sobolev OV Cryo-EM map interpretation and protein model-building using iterative map segmentation Protein Sci. 2020 29 87 99 10.1002/pro.3740 31599033 PMC6933853 · DOI
  14. [19]Johnson S Molecular structure of the intact bacterial flagellar basal body Nat. Microbiol. 2021 6 712 721 10.1038/s41564-021-00895-y 33931760 PMC7610862 · DOI
  15. [20]Kryshtafovych A Schwede T Topf M Fidelis K Moult J Critical assessment of methods of protein structure prediction (CASP)—round XIV Proteins Struct. Funct. Bioinf. 2021 89 1607 1617 10.1002/prot.26237 PMC8726744 34533838 · DOI · Paper page
  16. [21]Pozharski E Percentile-based spread: a more accurate way to compare crystallographic models Acta Crystallogr. Sect. D 2010 66 970 978 10.1107/S0907444910027927 20823548 · DOI
  17. [22]Pfab J Phan NM Si D DeepTracer for fast de novo cryo-EM protein structure modeling and special studies on CoV-related complexes Proc. Natl Acad. Sci. 2021 118 e2017525118 10.1073/pnas.2017525118 33361332 PMC7812826 · DOI
  18. [23]Afonine PV Real-space refinement in PHENIX for cryo-EM and crystallography Acta Crystallogr. Sect. D 2018 74 531 544 10.1107/S2059798318006551 PMC6096492 29872004 · DOI
  19. [24]Liebschner D Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix Acta Crystallogr. Sect. D 2019 75 861 877 10.1107/S2059798319011471 PMC6778852 31588918 · DOI
  20. [25]Emsley P Lohkamp B Scott WG Cowtan K Features and development of Coot Acta Crystallogr. Sect. D 2010 66 486 501 10.1107/S0907444910007493 20383002 PMC2852313 · DOI
  21. [26]Lawson CL EMDataBank.org: unified data resource for CryoEM Nucleic Acids Res. 2011 39 D456 D464 10.1093/nar/gkq880 20935055 PMC3013769 · DOI
  22. [27]Cragnolini T Kryshtafovych A Topf M Cryo-EM targets in CASP14 Proteins Struct. Funct. Bioinf. 2021 89 1949 1958 10.1002/prot.26216 PMC8630773 34398978 · DOI
  23. [28]Pettersen EF UCSF ChimeraX: structure visualization for researchers, educators, and developers Protein Sci. 2021 30 70 82 10.1002/pro.3943 32881101 PMC7737788 · DOI
  24. [29]Westbrook, J. D. & Fitzgerald, P. M. D. in Structural Bioinformatics Vol. 44 (eds Philip E. Bourne, P. E. & Weissig H.) Ch. 8 (John Wiley & Sons, 2003).
  25. [31]Cowtan K Metcalfe S Bond P Shift-field refinement of macromolecular atomic models Acta Crystallogr. Sect. D 2020 76 1192 1200 10.1107/S2059798320013170 PMC7709196 33263325 · DOI
  26. [32]Terwilliger TC Model morphing and sequence assignment after molecular replacement Acta Crystallogr. Sect. D 2013 69 2244 2250 10.1107/S0907444913017770 24189236 PMC3817698 · DOI
  27. [33]Kolbeck B May P Schmidt-Goenner T Steinke T Knapp E-W Connectivity independent protein-structure alignment: a hierarchical approach BMC Bioinf. 2006 7 510 10.1186/1471-2105-7-510 PMC1683948 17118190 · DOI

Cited by

Sorted by citations

Authors
Terwilliger, T. C., Poon, B. K., Afonine, P. V., Schlicksup, C. J., Croll, T. I., Millán, C., Richardson, J. S., Read, R. J., & Adams, P. D.
Affiliations
  • Los Alamos National Laboratory (US)
  • Lawrence Berkeley National Laboratory (US)
  • University of Cambridge (GB)
  • Duke University (US)
Corresponding author
Terwilliger, Thomas C.
ORCID
Venue
Nature Methods
Publisher
Springer Science and Business Media LLC
Published
2022-10-20
Volume
19
Issue
11
Pages
1376-1382
ISSN
  • 1548-7091 (ISSN-L)
  • 1548-7105
DOI
10.1038/s41592-022-01645-6
PMC
PMC9636017
OpenAlex
W4306888844
Cited by
140
References
6
Citations by year
2026: 11 · 2025: 35 · 2024: 29 · 2023: 53 · 2022: 12
FWCI
10.52
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Methods (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
  • eScholarship (California Digital Library) (submittedVersion, cc-by)
  • Apollo (University of Cambridge) (submittedVersion, cc-by)
  • Apollo
  • Apollo (cc-by)
Funders
  • U.S. Department of Energy (#89233218CNA000001, -AC02-05CH11231, DE-AC52-06NA2-5396, 05CH11231, DE AC02 05CH11231, AC52-06NA25396, DE-AC02-)
  • Wellcome Trust (209407/Z/17/Z)
  • National Institutes of Health (DE-AC02–05CH11231, DE-AC52-06NA25396, GM063210)
  • National Nuclear Security Administration (AC02-05CH11231, DE-AC02-05CH11231, DEAC52- 06NA25396, 89233218C-NA000001)
  • National Institute of General Medical Sciences (P01GM063210)
  • Lawrence Berkeley National Laboratory (AC02‐05CH11231, DE-AC02-05CH11231, 89233218CNA000001, 05CH11231)
  • Los Alamos National Laboratory (89233218CNA 000001, AC52-06NA25396, DE-AC52-06NA25396, DE-AC02-05CH11231)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
  • Physical Sciences › Materials Science › Materials Chemistry › Enzyme Structure and Function
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
Keywords
AlphaFold, protein structure prediction, model rebuilding, iterative refinement, template-based modeling, map interpretation, RoseTTAFold
MeSH terms
Machine Learning, Algorithms, Models, Molecular, Protein Conformation, Proteins, Cryoelectron Microscopy

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