CoChat

Limits and potential of combined folding and docking

Open access

Published version of a preprint

The DOI matches its Crossref record; 52 of the 52 references it cites were found in a registry or an index.

Abstract

Motivation: In the last decade, de novo protein structure prediction accuracy for individual proteins has improved significantly by utilising deep learning (DL) methods for harvesting the co-evolution information from large multiple sequence alignments (MSAs). The same approach can, in principle, also be used to extract information about evolutionary-based contacts across protein–protein interfaces. However, most earlier studies have not used the latest DL methods for inter-chain contact distance prediction. This article introduces a fold-and-dock method based on predicted residue-residue distances with trRosetta. Results: The method can simultaneously predict the tertiary and quaternary structure of a protein pair, even when the structures of the monomers are not known. The straightforward application of this method to a standard dataset for protein–protein docking yielded limited success. However, using alternative methods for generating MSAs allowed us to dock accurately significantly more proteins. We also introduced a novel scoring function, PconsDock, that accurately separates 98% of correctly and incorrectly folded and docked proteins. The average performance of the method is comparable to the use of traditional, template-based or ab initio shape-complementarity-only docking methods. Moreover, the results of conventional and fold-and-dock approaches are complementary, and thus a combined docking pipeline could increase overall docking success significantly. This methodology contributed to the best model for one of the CASP14 oligomeric targets, H1065. Availability and implementation: All scripts for predictions and analysis are available from https://github.com/ElofssonLab/bioinfo-toolbox/ and https://gitlab.com/ElofssonLab/benchmark5/. All models joined alignments, and evaluation results are available from the following figshare repository https://doi.org/10.6084/m9.figshare.14654886.v2. Supplementary information: Supplementary data are available at Bioinformatics online.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

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

Integrity signals

Registry

Crossref · 10.1093/bioinformatics/btab760

Cites

52 of 52 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1. Pipeline of the Fold-and-Dock protocol. First, two multiple sequence alignments are created using, for instance, jackhmmer. Second, these are merged by identifying potential pairs of orthologs. Then trRosetta is used to predict all pairwise distances. Here, the predicted (lower triangle) and actual (upper triangle) distance map for 4gmj is shown. The two blue stripes represent the poly-G linker between the two chains. Next, a model is created using pyConsfold,based on CNS (Chaudhury et al., 2010; Adams et al., 2013), or pyRosetta. Here, the natural (dark colours) and modeled (light colours) structure of the protein 4gmj is shown. The model’s accuracy is good, dockQ score 0.42, and the TM-scores for the two chains are 0.82 and 0.85, respectively. Finally, the models are scored using PconsDock, a consensus-based docking score. Here, two examples are shown 1vrs (left) where no consensus is seen and 1ay7 (right) where all five models are virtually identical
Fig. 1. Pipeline of the Fold-and-Dock protocol. First, two multiple sequence alignments are created using, for instance, jackhmmer. Second, these are merged by identifying potential pairs of orthologs. Then trRosetta is used to predict all pairwise distances. Here, the predicted (lower triangle) and actual (upper triangle) distance map for 4gmj is shown. The two blue stripes represent the poly-G linker between the two chains. Next, a model is created using pyConsfold,based on CNS (Chaudhury et al., 2010; Adams et al., 2013), or pyRosetta. Here, the natural (dark colours) and modeled (light colours) structure of the protein 4gmj is shown. The model’s accuracy is good, dockQ score 0.42, and the TM-scores for the two chains are 0.82 and 0.85, respectively. Finally, the models are scored using PconsDock, a consensus-based docking score. Here, two examples are shown 1vrs (left) where no consensus is seen and 1ay7 (right) where all five models are virtually identicalCC BY-NC · PubMed Central
Fig. 2. Performance of the fold-and-dock methodology versus the size of the joint alignments. Average TM-score of the two chains (A) and dockQ scores (B) plotted against the size of the multiple sequence alignment used to predict the contacts
Fig. 2. Performance of the fold-and-dock methodology versus the size of the joint alignments. Average TM-score of the two chains (A) and dockQ scores (B) plotted against the size of the multiple sequence alignment used to predict the contactsCC BY-NC · PubMed Central
Fig. 3. Results of Fold and Dock using different alignments. (A) dockQ scores for all models using six different alignments (see Supplementary Table S1). (B) The dockQ scores for the 15 proteins where at least one of these six alignments produce an acceptable model (dockQ > 0.23)
Fig. 3. Results of Fold and Dock using different alignments. (A) dockQ scores for all models using six different alignments (see Supplementary Table S1). (B) The dockQ scores for the 15 proteins where at least one of these six alignments produce an acceptable model (dockQ > 0.23)CC BY-NC · PubMed Central
Fig. 4. Results of alternative docking and folding protocols. Average dockQ score of the two chains (A) and TMscore scores (B). The default (N3) performance is compared with pyconsFold (uses the pyconsFold program instead of Rosetta), RaptorX (uses inter-chain contacts predicted by RaptorX instead of distances from trRosetta), RaptorX and N3-pdb use the intra-chain distances from the native structures, and N3-merged uses intra-chain distances predicted by the full alignments for each chain independently
Fig. 4. Results of alternative docking and folding protocols. Average dockQ score of the two chains (A) and TMscore scores (B). The default (N3) performance is compared with pyconsFold (uses the pyconsFold program instead of Rosetta), RaptorX (uses inter-chain contacts predicted by RaptorX instead of distances from trRosetta), RaptorX and N3-pdb use the intra-chain distances from the native structures, and N3-merged uses intra-chain distances predicted by the full alignments for each chain independentlyCC BY-NC · PubMed Central
Fig. 5. Features separating good (yellow) and bad (blue) models. Good models are defined as having a dockQ score over 0.23. All models from all the methods marked with a star in Supplementary Table S1 are included. (A) Distribution of the number of effective sequences (log10 of Meff), (B) average TM-scores of chain A and B, (C) number of inter-chain contacts predicted, (D) MCC values describing the accuracy of the inter-chain contacts, (E) number of contacts in the native interface and (F) HHprob (similarity) between chain a and chain b in the complex
Fig. 5. Features separating good (yellow) and bad (blue) models. Good models are defined as having a dockQ score over 0.23. All models from all the methods marked with a star in Supplementary Table S1 are included. (A) Distribution of the number of effective sequences (log10 of Meff), (B) average TM-scores of chain A and B, (C) number of inter-chain contacts predicted, (D) MCC values describing the accuracy of the inter-chain contacts, (E) number of contacts in the native interface and (F) HHprob (similarity) between chain a and chain b in the complexCC BY-NC · PubMed Central
Fig. 6. Distance maps for predicted (lower left) and native (upper right) distances for three proteins with artefacts 3qlu (A), 3pv6 (B) and 4yoc(C)
Fig. 6. Distance maps for predicted (lower left) and native (upper right) distances for three proteins with artefacts 3qlu (A), 3pv6 (B) and 4yoc(C)CC BY-NC · PubMed Central
Fig. 7. (A) Prediction qualities for models from different kingdoms of life using the N3-predictions. (B) Predictions qualities for all models using different alignments and sequence databases (see Supplementary Table S1 for details). (C) Comparison of dockQ scores for individual models compared to the N3 models
Fig. 7. (A) Prediction qualities for models from different kingdoms of life using the N3-predictions. (B) Predictions qualities for all models using different alignments and sequence databases (see Supplementary Table S1 for details). (C) Comparison of dockQ scores for individual models compared to the N3 modelsCC BY-NC · PubMed Central
Fig. 8. Comparison with GRAMM and TMdock. GRAMM-contact is the GRAMM-scores using predicted contacts as a constraint, GRAMM is GRAMM-docking ranked by the AACE18 scoring potential, TMdock-double is a template-based docking method
Fig. 8. Comparison with GRAMM and TMdock. GRAMM-contact is the GRAMM-scores using predicted contacts as a constraint, GRAMM is GRAMM-docking ranked by the AACE18 scoring potential, TMdock-double is a template-based docking methodCC BY-NC · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

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?

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

Cited by

See all 26

Metadata

Authors
Pozzati, G., Zhu, W., Bassot, C., Lamb, J., Kundrotas, P., & Elofsson, A.
Affiliations
  • Stockholm University (SE)
  • University of Kansas (US)
Corresponding author
Elofsson, Arne
Venue
Bioinformatics
Published
2021-11-12
Volume
38
Issue
4
Pages
954-961
ISSN
  • 1367-4803 (ISSN-L)
  • 1367-4811
OpenAlex
W3212532687
Cited by
25
References
52
Citations by year
2026: 2 · 2025: 1 · 2024: 8 · 2023: 6 · 2022: 6 · 2021: 2
FWCI
1.73
Type
journal article
Language
en
Open access
Locations
  • Bioinformatics (publishedVersion, cc-by-nc)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
Funders
  • Swedish National Research Council (VR-NT-2016-03798)
  • Swedish Research Council partly paid the salary
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
protein-protein docking, de novo protein structure prediction, multiple sequence alignment, quaternary structure prediction, oligomeric proteins
MeSH terms
Algorithms, Proteins, Sequence Alignment, Computational Biology

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

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.