EUVIMEDObjevte EUVIMED
Open access · CC BY 4.0

Assessing scoring metrics for AlphaFold2 and AlphaFold3 protein complex predictions

Luca R Genz, Sanjana Nair, Natan Nagar, Maya Topf iD

Protein Science · 2025-11 · 34 (11)

DOI: 10.1002/pro.70327 ↗

At a glance

Journal article in Protein Science, published online on 2025-10-13. Written by Luca R Genz and 3 co-authors. Institution: Research Department of Integrative Virology, Leibniz-Institut für Virologie (LIV), Germany. Major MEDLINE topics: Proteins, Models, Molecular, Software, Computational Biology. Cited 32 times so far (Crossref); 40 references. The full text is freely available (open access, CC BY 4.0). Funded by: Deutsche Forschungsgemeinschaft, Bundesministerium für Bildung und Forschung.

AbstraktSubjectsBibliografické údajeCiteRelated studies

Abstract (original language)

Recent breakthroughs in AI‐driven protein structure prediction have revolutionized structural biology, unlocking new possibilities to model complex biomolecular interactions. We evaluated widely used scoring metrics for assessing models predicted by ColabFold with templates, ColabFold without templates, and AlphaFold3. We benchmarked the optimal cutoffs for these assessment scores using a set of 223 heterodimeric, high‐resolution protein structures and their predictions. Our results show that ColabFold with templates and AlphaFold3 perform similarly, and both outperform ColabFold without templates. However, the assessment scores perform best on ColabFold without templates. Furthermore, interface‐specific scores are more reliable for evaluating protein complex predictions compared to the corresponding global scores. Notably, ipTM and model confidence achieve the best discrimination between correct and incorrect predictions. Based on our results, we developed a weighted combined score, C2Qscore, to improve model quality assessment. We used C2Qscore to analyze dimers from large assemblies solved by cryoEM, revealing potential limitations of the existing metrics when multiple configurations of heterodimers are possible. This study provides insights into the strengths and weaknesses of current scores and offers guidance for improving protein complex model assessment under realistic use case conditions. C2Qscore has been integrated as a tool into our ChimeraX plug‐in PICKLUSTER v.2.0 and is also available as a command‐line tool on https://gitlab.com/topf-lab/c2qscore .

Subjects

MeSH terms (MEDLINE)

ProteinsModels, MolecularSoftwareComputational BiologyProtein ConformationDatabases, ProteinProtein MultimerizationProtein Folding

Author keywords

AlphaFold3ChimeraX plug‐inColabFoldInterface scoringProtein complex modelingProtein structure predictionScoring metrics

Substances

Proteins

Authors and institutions

  • Research Department of Integrative Virology, Leibniz-Institut für Virologie (LIV), Germany.
  • Department of Chemistry, Universität Hamburg, Germany.
  • Centre for Structural Systems Biology (CSSB), Hamburg, Germany.
  • Viral Data Science, Leibniz-Institut für Virologie (LIV), Germany.
  • Institute for Molecular Virology and Tumorvirology, Universitätsklinikum Hamburg Eppendorf (UKE), Germany.

ORCID profiles: Maya Topf

Funding

  • Deutsche Forschungsgemeinschaft
  • Bundesministerium für Bildung und Forschung

Cite

APA

Genz, L. R., Nair, S., Nagar, N., & Topf, M. (2025). Assessing scoring metrics for AlphaFold2 and AlphaFold3 protein complex predictions. Protein Science, 34(11). https://doi.org/10.1002/pro.70327

Vancouver

Genz LR, Nair S, Nagar N, Topf M. Assessing scoring metrics for AlphaFold2 and AlphaFold3 protein complex predictions. Protein Science. 2025;34(11). doi:10.1002/pro.70327

BibTeX
@article{Genz2025,
  author = {Genz, Luca R and Nair, Sanjana and Nagar, Natan and Topf, Maya},
  title = {Assessing scoring metrics for AlphaFold2 and AlphaFold3 protein complex predictions},
  journal = {Protein Science},
  year = {2025},
  volume = {34},
  number = {11},
  doi = {10.1002/pro.70327}
}
Start for free →