CoChat

QSProteome: A Community-Driven Interactive Platform for Large-Scale Exploration and Evaluation of Predicted Protein Complex Structures

Open access
Catoiu, Edward A., Kambo, Delina, Rodriguez, Brianna, Vear, Krista, Wang, Ziheng, Akpata, June, Shah, Jaladhi, Ayoub, Rayyan, Hofmann, Matthew, Ananda, Keerthana, Tantry, Avani, Gupta, Ishita, Anwyl, Robin, Faucher, Carter, Nichols, Jordan, Ton, Brian, Marchal, Lorenzo O., Palsson, Bernhard O. · 2025 · Nucleic Acids Research

Preprint — published as 10.1093/nar/gkaf1099

The DOI matches its Crossref record; 2 of the 2 references it cites were found in a registry or an index.

Abstract

QSProteome ( https://QSProteome.org ) is a community-scale platform for modeling, evaluating, and refining quaternary protein structure. The resource hosts 35,528 unique modeled assemblies spanning over 42,000 genes, covering nearly all curated complexes in BioCyc and ComplexPortal databases. Each model is displayed on an interactive page with 3D visualization, chain-level confidence metrics, structural alignments, functional annotations, and automated stoichiometry checks against database expectations. A cloud-based server supports continuous user uploads and automated processing pipelines—enabling submission and validation of >54,000 models within 14 weeks. To promote iterative refinement, QSProteome includes a gamified re-curation workflow that transitions users from training modules into live curation, enabling the community-led assessment of 1,547 ABC transporter complexes. Together, these components form a dynamic, scalable infrastructure for proteome-scale structural biology. By unifying modeling, validation, and annotation in a reusable, searchable, and community-extensible framework, QSProteome enables proteome-scale structure accessibility and reuse—powering discovery, annotation, and collaborative refinement across the structural biology community.

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 · 10.1101/2025.09.10.675416

Cites

2 of 2 references verified0 citation errors · 0 not checkable

Statistics

Figures

Graphical Abstract
Graphical Abstract CC BY-NC · PubMed Central
Figure 1. The QSProteome Upload Portal (A) guides users through: (i) entering optional personal details for attribution, (ii) uploading a zip archive of up to 30 models, (iii) previewing each model, screenshotting (for search result rendering) and uploading, and (iv) tracking the status of uploaded models. (B) QSProteome automatically: (1) detects configuration, metrics, and coordinate files; (2) checks for duplicate uploads; (3) validates homo-oligomeric models via QSalignWeb; (4) retrieves sequence- and gene-level data from UniParc/UniProt [22]; (5) conducts structural alignments [24] between the model and related structures in RCSB PDB [23]; and (6) maps models to external database entries. A screenshot of each model is stored for display on the Search Page. QSalign and MMalign aggregate results are displayed on the Assessment Page. Retrieved data is cached in server-side databases to minimize future queries to third-party APIs.
Figure 1. The QSProteome Upload Portal (A) guides users through: (i) entering optional personal details for attribution, (ii) uploading a zip archive of up to 30 models, (iii) previewing each model, screenshotting (for search result rendering) and uploading, and (iv) tracking the status of uploaded models. (B) QSProteome automatically: (1) detects configuration, metrics, and coordinate files; (2) checks for duplicate uploads; (3) validates homo-oligomeric models via QSalignWeb; (4) retrieves sequence- and gene-level data from UniParc/UniProt [22]; (5) conducts structural alignments [24] between the model and related structures in RCSB PDB [23]; and (6) maps models to external database entries. A screenshot of each model is stored for display on the Search Page. QSalign and MMalign aggregate results are displayed on the Assessment Page. Retrieved data is cached in server-side databases to minimize future queries to third-party APIs.CC BY-NC · PubMed Central
Figure 2. The size of the QSProtome Repository. (A) QSProteome hosts 35 528 unique structural models across 153 species-level taxa. Each unique gene-stoichiometry combination is shown once. (B) Gene signature overlap between QSProteome and annotated protein complexes. Proteins larger than 5000 amino acids or containing ambiguous sequence characters fail to produce a valid AlphaFold 3 structure. (C) The Repository page summarizes modeling progress, displaying completed versus target counts and color-coded AFM (ranking) scores (see Supplementary methods 5.9.3 of Abramson et al. 2024 [16]). Users can bulk-download pre-formatted AlphaFoldServer inputs, models, and metadata to streamline batch submission. New projects encourage community participation. Only highest-scoring unique models are shown. Disorder and has_clash are not used in AFM score calculation for models generated with AlphaFold v2 (scoring metric defined by Evans et al. 2022 [5]).
Figure 2. The size of the QSProtome Repository. (A) QSProteome hosts 35 528 unique structural models across 153 species-level taxa. Each unique gene-stoichiometry combination is shown once. (B) Gene signature overlap between QSProteome and annotated protein complexes. Proteins larger than 5000 amino acids or containing ambiguous sequence characters fail to produce a valid AlphaFold 3 structure. (C) The Repository page summarizes modeling progress, displaying completed versus target counts and color-coded AFM (ranking) scores (see Supplementary methods 5.9.3 of Abramson et al. 2024 [16]). Users can bulk-download pre-formatted AlphaFoldServer inputs, models, and metadata to streamline batch submission. New projects encourage community participation. Only highest-scoring unique models are shown. Disorder and has_clash are not used in AFM score calculation for models generated with AlphaFold v2 (scoring metric defined by Evans et al. 2022 [5]).CC BY-NC · PubMed Central
Figure 3. QSProteome Search and Retrieval Interface. (A) The keyword search retrieves models or known assemblies using gene names, protein IDs, functions, and ligands. (B) Advanced Search enables refined filtering by organism, gene name, model availability, complex size, confidence scoring, and oligomerization type. (C) Each result includes an image of the structure, AlphaFold metrics, QSalign validation (if available), MM-align alignments, complex size, gene stoichiometry, cross-referenced annotations, and ligands. Protein images link to dedicated model pages. All linked IDs (UniProt, ChEBI [26], ComplexPortal, BioCyc, PubMed) redirect externally. Colored badges indicate stoichiometric matches; e.g. CPX-4465 includes an extra copy of ynjC. For unmatched complexes, downloadable AlphaFold 3 input files are available. (D) Results are exportable in JSON and CSV formats.
Figure 3. QSProteome Search and Retrieval Interface. (A) The keyword search retrieves models or known assemblies using gene names, protein IDs, functions, and ligands. (B) Advanced Search enables refined filtering by organism, gene name, model availability, complex size, confidence scoring, and oligomerization type. (C) Each result includes an image of the structure, AlphaFold metrics, QSalign validation (if available), MM-align alignments, complex size, gene stoichiometry, cross-referenced annotations, and ligands. Protein images link to dedicated model pages. All linked IDs (UniProt, ChEBI [26], ComplexPortal, BioCyc, PubMed) redirect externally. Colored badges indicate stoichiometric matches; e.g. CPX-4465 includes an extra copy of ynjC. For unmatched complexes, downloadable AlphaFold 3 input files are available. (D) Results are exportable in JSON and CSV formats.CC BY-NC · PubMed Central
Figure 4. Transparent Model Assessment and Validation. (A) The embedded Mol* viewer allows interactive structure visualization with both standard and pLDDT-based color schemes. (B) AlphaFold metrics (ipTM, pTM, pLDDT, PAE, chain min PAE, and ipSAE) are displayed alongside modeling parameters. (C) QSalign validation results for homo-oligomers report quaternary structure conservation, symmetry class, subunit count, and confidence scores. (D) Users can overlay their model with a QSalign-validated target to visually assess structural similarity. Similar functionality is enabled for related PDB entries (not shown).
Figure 4. Transparent Model Assessment and Validation. (A) The embedded Mol* viewer allows interactive structure visualization with both standard and pLDDT-based color schemes. (B) AlphaFold metrics (ipTM, pTM, pLDDT, PAE, chain min PAE, and ipSAE) are displayed alongside modeling parameters. (C) QSalign validation results for homo-oligomers report quaternary structure conservation, symmetry class, subunit count, and confidence scores. (D) Users can overlay their model with a QSalign-validated target to visually assess structural similarity. Similar functionality is enabled for related PDB entries (not shown).CC BY-NC · PubMed Central
Figure 5. Integrating Biological Context into Structural Models. (A) The sequence viewer highlights selected residues on the 3D structure. (B) UniProt sequence features are displayed and linked to the structure viewer. (C) A STRING interaction network visualizes evidence of physical association among modeled subunits. (D) ComplexPortal annotations are shown for matching assemblies, including subunit composition and functional descriptions.
Figure 5. Integrating Biological Context into Structural Models. (A) The sequence viewer highlights selected residues on the 3D structure. (B) UniProt sequence features are displayed and linked to the structure viewer. (C) A STRING interaction network visualizes evidence of physical association among modeled subunits. (D) ComplexPortal annotations are shown for matching assemblies, including subunit composition and functional descriptions.CC BY-NC · PubMed Central
Figure 6. Model-guided refinement of ABC transporter assemblies through gamified annotation and automated curation. (A) The ABC Transporter Game guides users through model assessment tasks in a structured and engaging format. A progress bar tracks difficulty and performance (top), while a Mol* viewer (left) allows users to interact with the prompted model. Users assess whether the model resembles an ABC transporter and annotate subunit composition and class (center). Canonical examples provide visual reference for transporter types (bottom left), and feedback (including structural alignment to the correct canonical structure) is displayed after answer submission (right), reinforcing accurate identification and classification. (B) Initial ABC models were generated based on gene stoichiometries from public databases. The visual inspection and gene content of each model informed its classification for either automated or manual re-curation. (C) Minor stoichiometric issues were corrected via name parsing; major domain omissions triggered manual updates using operon context. (D) Amino acid length increased after curation, reflecting restoration of missing domains. (E) Inter-chain TM-scores (ipTM) improved post-curation, approaching confidence levels of known high-quality assemblies (*paired Wilcoxon, ✝Mann–Whitney).
Figure 6. Model-guided refinement of ABC transporter assemblies through gamified annotation and automated curation. (A) The ABC Transporter Game guides users through model assessment tasks in a structured and engaging format. A progress bar tracks difficulty and performance (top), while a Mol* viewer (left) allows users to interact with the prompted model. Users assess whether the model resembles an ABC transporter and annotate subunit composition and class (center). Canonical examples provide visual reference for transporter types (bottom left), and feedback (including structural alignment to the correct canonical structure) is displayed after answer submission (right), reinforcing accurate identification and classification. (B) Initial ABC models were generated based on gene stoichiometries from public databases. The visual inspection and gene content of each model informed its classification for either automated or manual re-curation. (C) Minor stoichiometric issues were corrected via name parsing; major domain omissions triggered manual updates using operon context. (D) Amino acid length increased after curation, reflecting restoration of missing domains. (E) Inter-chain TM-scores (ipTM) improved post-curation, approaching confidence levels of known high-quality assemblies (*paired Wilcoxon, ✝Mann–Whitney).CC 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

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

Show all 2 references

Research works printed but not linked: 44

  1. 10.1038/s41592-022-01488-1 · doi:10.1038/s41592-022-01488-1 · Paper page
  2. 10.1042/BCJ20200963 · DOI · Paper page
  3. 10.1002/prot.26237 · DOI · Paper page
  4. 10.1101/2021.10.04.463034 · DOI · Paper page
  5. 10.1016/j.jmb.2025.168996 · DOI · Paper page
  6. 10.1038/nmeth.4510 · DOI
  7. 10.1016/j.str.2021.07.012 · DOI
  8. 10.3389/fmolb.2021.787510 · DOI
  9. 10.1002/pmic.202200323 · DOI · Paper page
  10. 10.1016/j.cell.2024.01.022 · DOI · Paper page
  11. 10.1093/nar/gkae1010 · DOI · Paper page
  12. 10.1093/nar/gkab991 · DOI · Paper page
  13. 10.1093/nar/gkac1000 · DOI · Paper page
  14. 10.1038/s41586-024-07487-w · DOI · Paper page
  15. 10.1093/bib/bbx085 · DOI · Paper page
  16. 10.1128/ecosalplus.ESP-0002-2023 · DOI
  17. 10.1093/nar/gkz862 · DOI · Paper page
  18. 10.7554/eLife.100485.1 · DOI
  19. 10.1016/j.jmb.2021.167351 · DOI · Paper page
  20. 10.1093/nar/gkaf394 · DOI · Paper page
  21. 10.1093/nar/gkae1091 · DOI · Paper page
  22. 10.1093/nar/gkp318 · DOI · Paper page
  23. 10.1038/s41597-024-03571-y · DOI · Paper page
  24. 10.1093/nar/gkv1031 · DOI · Paper page
  25. 10.1093/nar/gkab314 · DOI · Paper page
  26. 10.1101/2025.02.10.637595 · DOI · Paper page
  27. 10.1038/nrm2646 · DOI · Paper page
  28. 10.1128/msystems · DOI
  29. 10.3389/fmicb.2024.1469915 · DOI
  30. 10.1039/C5RA05304H · DOI
  31. 10.1016/j.febslet.2010.04.043 · DOI · Paper page
  32. 10.1038/s41587-023-01773-0 · DOI · Paper page
  33. 10.1007/978-3-030-78397-6_2 · DOI
  34. 10.1146/annurev.micro.091208.073214 · DOI · Paper page
  35. 10.1038/s41579-023-00974-3 · DOI · Paper page
  36. 10.4161/rna.24538 · DOI
  37. 10.1002/prot.26617 · DOI · Paper page
  38. 10.1093/bioinformatics/btaf107 · DOI · Paper page
  39. 10.1093/bioinformatics/btx283 · DOI
  40. 10.1038/s41587-024-02175-6 · DOI · Paper page
  41. 10.1371/journal.pcbi.1006302 · DOI · Paper page
  42. 10.1073/pnas.1705524114 · DOI · Paper page
  43. 10.1371/journal.pcbi.1011865 · DOI
  44. 10.1038/nbt.3597 · DOI · Paper page

Metadata

Authors
Catoiu, E. A., Kambo, D., Rodriguez, B., Vear, K., Wang, Z., Akpata, J., Shah, J., Ayoub, R., Hofmann, M., Ananda, K., Tantry, A., Gupta, I., Anwyl, R., Faucher, C., Nichols, J., Ton, B., Marchal, L. O., & Palsson, B. O.
Affiliations
University of California San Diego (US)
Corresponding author
Catoiu, Edward A.
Venue
Nucleic Acids Research
Publisher
openRxiv
Published
2025-09-16
Volume
54
Issue
D1
Pages
D661-D672
ISSN
0305-1048 (ISSN-L)
OpenAlex
W4415721619
Cited by
4
References
2
Citations by year
2026: 3 · 2025: 1
FWCI
1.20
Type
journal article
Language
en
Open access
Locations
  • Nucleic Acids Research (publishedVersion, cc-by-nc)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, other-oa)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed Central (submittedVersion)
Funders
  • Novo Nordisk Foundation (NNF20CC0035580)
  • National Institutes of Health (R01 GM057089)
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 › Bioinformatics and Genomic Networks
Keywords
protein complex structure prediction, structural alignment, 3D visualization

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

How to cite

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

This page
CoChat. (2026). Citation check: QSProteome: A Community-Driven Interactive Platform for Large-Scale Exploration and Evaluation of Predicted Protein Complex Structures. https://app.cochat.ai/papers/10.1101/2025.09.10.675416
The paper
Catoiu, E. A., Kambo, D., Rodriguez, B., Vear, K., Wang, Z., Akpata, J., Shah, J., Ayoub, R., Hofmann, M., Ananda, K., Tantry, A., Gupta, I., Anwyl, R., Faucher, C., Nichols, J., Ton, B., Marchal, L. O., & Palsson, B. O. (2025). QSProteome: A Community-Driven Interactive Platform for Large-Scale Exploration and Evaluation of Predicted Protein Complex Structures. Nucleic Acids Research. https://doi.org/10.1101/2025.09.10.675416

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.