CoChat

Using AlphaFold to predict the impact of single mutations on protein stability and function

Open access

Published version of a preprint

The DOI matches its Crossref record; 33 of the 33 references it cites were found in a registry or an index; 2 of the 9 sources its claims cite were checked for support.

Abstract

AlphaFold changed the field of structural biology by achieving three-dimensional (3D) structure prediction from protein sequence at experimental quality. The astounding success even led to claims that the protein folding problem is “solved”. However, protein folding problem is more than just structure prediction from sequence. Presently, it is unknown if the AlphaFold-triggered revolution could help to solve other problems related to protein folding. Here we assay the ability of AlphaFold to predict the impact of single mutations on protein stability (ΔΔG) and function. To study the question we extracted the pLDDT and metrics from AlphaFold predictions before and after single mutation in a protein and correlated the predicted change with the experimentally known ΔΔG values. Additionally, we correlated the same AlphaFold pLDDT metrics with the impact of a single mutation on structure using a large scale dataset of single mutations in GFP with the experimentally assayed levels of fluorescence. We found a very weak or no correlation between AlphaFold output metrics and change of protein stability or fluorescence. Our results imply that AlphaFold may not be immediately applied to other problems or applications in protein folding.

Overview

CoChat Summary

TL;DR

The paper tests whether AlphaFold confidence metrics can predict the effect of single mutations on protein stability (ΔΔG) and GFP fluorescence, using experimental mutation datasets. It finds weak or no useful correlation, concluding AlphaFold outputs are not directly reliable for these prediction tasks.

Key findings

  • For 1154 mutations in 73 proteins, the correlation between mutation-induced change in local confidence and experimental stability change was weak: PCC = -0.17 ± 0.03 for ΔpLDDT, while Δ<pLDDT> was 0.02 ± 0.03.
  • A multiple linear regression combining ΔpLDDT and Δ<pLDDT> did not improve prediction much: 0.12 ± 0.05 on the training set and 0.20 ± 0.04 on the test set.
  • For 796 GFP single mutants, AlphaFold metrics also showed only weak association with function: PCC = 0.17 ± 0.03 for ΔpLDDT and 0.16 ± 0.04 for Δ<pLDDT>.
  • Among destabilizing mutations, 87% had negative ΔpLDDT, but this still did not produce a meaningful correlation with ΔΔG.

Methods

  • Experimental stability data came from ThermoMutDB v1.3; after filtering for single-point mutations measured at pH 3–9, 293–300 K, and proteins <250 aa, the authors analyzed 1154 mutations in 73 proteins.
  • GFP function data came from a published dataset; the authors randomly selected 796 single mutants and used fluorescence as the functional readout.
  • Wild-type and mutant structures were obtained from the AlphaFold Protein Structure Database or predicted with standalone AlphaFold, and analysis focused on ΔpLDDT and Δ<pLDDT>; the stability dataset was split into training/test sets by a 50% sequence identity threshold and analyzed with multiple linear regression.

Limitations

  • The paper states that AlphaFold was not designed or validated to predict mutation effects on stability or function, so direct use is not expected to work well.
  • The authors note that the available ΔΔG data are limited in size (they describe experimentally measured ΔΔG values as being on the order of 10,000), which they argue constrains deep-learning prediction; this is an interpretation made in the paper.
  • The evidence is limited to the datasets studied here: single-point mutations in ThermoMutDB proteins and GFP mutants, so broader generalization is not stated in the paper.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0282689

Cites

33 of 33 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig 1 Correlation between ΔΔG and ΔpLDDT. Correlation between the effect of mutation on protein stability, ΔΔG, and change of confidence score of structure prediction, ΔpLDDT. A: The correlation for the mutated amino acid. B: The correlation for the whole structure.
Fig 1 Correlation between ΔΔG and ΔpLDDT. Correlation between the effect of mutation on protein stability, ΔΔG, and change of confidence score of structure prediction, ΔpLDDT. A: The correlation for the mutated amino acid. B: The correlation for the whole structure.CC BY · PubMed Central
Fig 2 Correlation between the GFP fluorescence and ΔpLDDT. Correlation between the GFP fluorescence and change of confidence score of structure prediction, ΔpLDDT. A: The correlation for the mutated amino acid. B: The correlation for the whole structure.
Fig 2 Correlation between the GFP fluorescence and ΔpLDDT. Correlation between the GFP fluorescence and change of confidence score of structure prediction, ΔpLDDT. A: The correlation for the mutated amino acid. B: The correlation for the whole structure.CC BY · PubMed Central

Verification

Claims the paper makes

2 Valid · 0 Not valid · 4 Not checked · 19 Own results and interpretations

14 more claims

Create a free account to see all 25 claims and to verify the 2 references still open.

See all claims — free account
  • Probably not a substantial claim. 2
  • It may state two separate claims as one. 4

Create a free account to see these claims.

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?

Citations

Cites

Show all 33 references

Cited by

See all 322

Metadata

Authors
Pak, M. A., Markhieva, K. A., Novikova, M. S., Petrov, D. S., Vorobyev, I. S., Maksimova, E. S., Kondrashov, F. A., & Ivankov, D. N.
Affiliations
  • Skolkovo Institute of Science and Technology (RU)
  • Peoples' Friendship University of Russia (RU)
  • University of New Mexico (US)
  • Ural Federal University (RU)
  • Institute of Science and Technology Austria (AT)
  • Okinawa Institute of Science and Technology Graduate University (JP)
Corresponding author
Ivankov, Dmitry N.
Venue
PLOS One
Published
2023-03-16
Volume
18
Issue
3
Pages
e0282689
ISSN
1932-6203 (ISSN-L)
OpenAlex
W4327566741
Cited by
322
References
32
Citations by year
2026: 51 · 2025: 90 · 2024: 77 · 2023: 66
FWCI
25.19
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Electronic Archive of Ural Federal University (ELAR UrFU) (submittedVersion, cc-by)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Structural Biology › Enzyme Structure and Function
Keywords
AlphaFold, single mutation, protein stability, protein folding, pLDDT, ΔΔG, protein function, green fluorescent protein, fluorescence, protein structure prediction, protein folding problem
MeSH terms
Amino Acid Sequence, Mutation, Proteins, Protein Folding, Protein Stability

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.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Using AlphaFold to predict the impact of single mutations on protein stability and function. https://app.cochat.ai/papers/10.1371/journal.pone.0282689
The paper
Pak, M. A., Markhieva, K. A., Novikova, M. S., Petrov, D. S., Vorobyev, I. S., Maksimova, E. S., Kondrashov, F. A., & Ivankov, D. N. (2023). Using AlphaFold to predict the impact of single mutations on protein stability and function. PLOS One. https://doi.org/10.1371/journal.pone.0282689

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.