EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Large-scale prospective evaluation of co-folding across hundreds of Mac1–ligand complexes and three virtual screens

Jongbin Kim iD, Galen J Correy iD, Brendan W Hall, Moira M Rachman, Olivier Mailhot iD, Takaya Togo iD, Ryan L Gonciarz iD, Priyadarshini Jaishankar iD, R Jeffrey Neitz iD, Eric R Hantz iD, Yagmur U Doruk, Maisie G V Stevens iD und 8 weitere

eLife · 12.10.2026 · 15

DOI: 10.7554/elife.110475 ↗

Kurz gefasst

Fachartikel in der Zeitschrift eLife, online erschienen am 12.10.2026. Verfasst von Jongbin Kim und 19 weiteren Autor:innen. Beteiligte Einrichtung: Department of Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, United States. Hauptthemen laut MEDLINE: Protein Folding, Viral Nonstructural Proteins. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 75 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NIGMS NIH HHS, NIH HHS, NIAID NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate prediction of ligand-bound protein complexes and ranking them by affinity are central problems in drug discovery. While deep-learning co-folding methods can help address these challenges, their evaluation has been hampered by the difficulties in assessing independence from training data and insufficiently large test sets. Here, we test the ability of co-folding methods to predict the structures of 551 ligands, 489 of which are of sufficient quality for evaluation, bound to the SARS-CoV-2 NSP3 macrodomain (Mac1) that were determined after the training cutoff dates. AlphaFold3 (AF3), Boltz-2, and Chai-1 each reproduced >50% of the Mac1 ligand poses to better than 2 Å root-mean-square deviation of experiment. Despite the potential for co-folding to describe protein conformational changes that stabilize ligand binding, we did not find that common conformational rearrangements, including peptide flip and a large loop opening, were recapitulated by the co-folding prediction. For AF3 and Chai-1, ligand pose prediction confidence weakly, but significantly, tracked experimental potency, while DOCK3.7 energies were only weakly correlated. Boltz-2 affinity predictions showed the strongest correlation with measured potency and, after calibration, achieved lower mean absolute error than a baseline predictor. We next assessed whether co-folding scores could rescore docking hit lists to distinguish true ligands from non-binders among hundreds of molecules prospectively experimentally tested against AmpC β-lactamase, the dopamine D4, and the σ 2 receptors. AF3 ligand pose confidence values did not separate true ligands from high-scoring false positives as effectively as docking scores or Boltz-2 affinity predictions did. Taken together, the modest, but independent correlations of docking score and co-folding confidence or affinity suggest that integrating physics-based and deep-learning approaches may help with hit prioritization and subsequent optimization in structure-based ligand discovery.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Protein FoldingViral Nonstructural ProteinsLigandsProtein BindingSARS-CoV-2Protein ConformationHumansDrug Discovery

Stichwörter der Autor:innen

E. coliX-ray crystallographyco-foldingdockingdrug designmachine learningmolecular biophysicsstructural biologystructure prediction

Substanzen

LigandsViral Nonstructural ProteinsORF1ab polyprotein, SARS-CoV-2

Autor:innen und Institutionen

  • Department of Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, United States.
  • Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, United States.
  • Small Molecule Discovery Center, Department of Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, United States.
  • Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, United States.
  • Department of Biological and Medical Informatics, University of California, San Francisco, San Francisco, United States.
  • Quantitative Biosciences Institute (QBI), University of California, San Francisco, San Francisco, United States.

ORCID-Profile: Jongbin Kim, Galen J Correy, Olivier Mailhot, Takaya Togo, Ryan L Gonciarz, Priyadarshini Jaishankar, R Jeffrey Neitz, Eric R Hantz, Maisie G V Stevens, Morgan E Diolaiti, Rashad Reid, Saumya Gopalkrishnan, Brian K Shoichet, James S Fraser

Förderung

  • NIGMS NIH HHS
  • NIH HHS
  • NIAID NIH HHS
  • National Institutes of Health

Zitieren

APA

Kim, J., Correy, G. J., Hall, B. W., Rachman, M. M., Mailhot, O., Togo, T., Gonciarz, R. L., Jaishankar, P., Neitz, R. J., Hantz, E. R., Doruk, Y. U., Stevens, M. G. V., Diolaiti, M. E., Reid, R., Gopalkrishnan, S., Krogan, N. J., Renslo, A. R., Ashworth, A., Shoichet, B. K., & Fraser, J. S. (2026). Large-scale prospective evaluation of co-folding across hundreds of Mac1–ligand complexes and three virtual screens. eLife, 15. https://doi.org/10.7554/elife.110475

Vancouver

Kim J, Correy GJ, Hall BW, Rachman MM, Mailhot O, Togo T, et al. Large-scale prospective evaluation of co-folding across hundreds of Mac1–ligand complexes and three virtual screens. eLife. 2026;15. doi:10.7554/elife.110475

BibTeX
@article{Kim2026,
  author = {Kim, Jongbin and Correy, Galen J and Hall, Brendan W and Rachman, Moira M and Mailhot, Olivier and Togo, Takaya and Gonciarz, Ryan L and Jaishankar, Priyadarshini and Neitz, R Jeffrey and Hantz, Eric R and Doruk, Yagmur U and Stevens, Maisie G V and Diolaiti, Morgan E and Reid, Rashad and Gopalkrishnan, Saumya and Krogan, Nevan J and Renslo, Adam R and Ashworth, Alan and Shoichet, Brian K and Fraser, James S},
  title = {Large-scale prospective evaluation of co-folding across hundreds of Mac1–ligand complexes and three virtual screens},
  journal = {eLife},
  year = {2026},
  volume = {15},
  doi = {10.7554/elife.110475}
}
Kostenlos starten →