Datasets
SAbDab antibody–antigen complexes and the dMaSIF protein–protein interaction dataset.
Task
Antibody–antigen interaction prediction using folded complexes
Antibody–antigen interaction evaluation uses a curated SAbDab collection alongside a separate protein-interaction dataset.
Sources
Enhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passagesOverview
Metrics
ROC-AUC for the interaction task, comparing original and folded complex inputs; paratope/epitope metrics are separate evaluations.
Allowed inputs
Folded antibody–antigen complexes.
Sources
Enhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passagesEvaluation procedure diagram
How it worksComputational evaluation flow
Conceptual summary of the cited evaluation; exact task configuration and source version remain part of the protocol.
Sources
Enhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passageslimited source coverage · Automated source review, 2026-09-16. All specifications and missing details
Results
Results are available, but no reviewed comparison panel is linked in this release.
All evaluations
2 evaluations · 2 results. Different protocols are not a single leaderboard.
Filter evaluations
Applied filters: All linked evaluations
| Tested configuration | Protocol and dataset | Finding | Evidence and details |
|---|---|---|---|
| Configuration: Chai-1 | Task: Antibody–antigen interaction prediction using folded complexes Dataset: Antibody–antigen GEP test set | 0.86 AUC-ROC unitless · unknown Uncertainty: ± 0.07 Coverage: Not reported scored / Not reported eligible | Independent external evaluation · Source checkedMethods, coverage and sourceChai-1: Antibody–antigen interaction prediction using folded complexes Interaction classifier evaluated using Chai-1-folded input complexes; this is pipeline AUC, not DockQ. Aggregation: Not reported Enhancing antibody-antigen interaction prediction with atomic flexibility · Table 5, Folded row, Chai-1 (no MSA) column |
| Configuration: Boltz-1 | Task: Antibody–antigen interaction prediction using folded complexes Dataset: Antibody–antigen GEP test set | 0.85 AUC-ROC unitless · unknown Uncertainty: ± 0.05 Coverage: Not reported scored / Not reported eligible | Independent external evaluation · Source checkedMethods, coverage and sourceBoltz-1: Antibody–antigen interaction prediction using folded complexes Interaction classifier evaluated using Boltz-1-folded input complexes; this is pipeline AUC, not DockQ. Aggregation: Not reported Enhancing antibody-antigen interaction prediction with atomic flexibility · Table 5, Folded row, Boltz-1 (no MSA) column |
Source checking is not independent reproduction. Release 2026-09-29-06401fd5b220.
Methods and evaluation design
Procedure, tasks and evaluated configurations
How it works
Evaluation methodology
SAbDab antibody–antigen complexes and the dMaSIF protein–protein interaction dataset. The antibody–antigen collection is randomly partitioned with a validation subset and separate test complexes. ROC-AUC for the interaction task, comparing original and folded complex inputs; paratope/epitope metrics are separate evaluations. The interaction predictor is evaluated on original structures and complexes folded by AlphaFold3, Chai-1 and Boltz-1; these are alternative input-structure conditions. The source describes sequence deduplication and structural-similarity filtering before the random split.
Sources
Enhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passagesRecorded evaluations
Each evaluation records what was tested and under which conditions.
Run instructions
No runnable recipe has been reviewed for this task. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.
A task describes a biological question. Choose a linked protocol to obtain concrete split and scoring instructions.
Strengths, limitations and unresolved questions
Strengths and limitations
Strengths and considerations
No source-reviewed explanatory claims are recorded here yet.
Limitations and conditions
- The paratope/epitope classification and folded-complex interaction evaluations have different outcomes. Their thresholds and splits cannot be substituted for one another; predicted structure confidence is not benchmark uncertainty.
Sources
Enhancing antibody-antigen interaction prediction with atomic flexibility · Tables 3–5 captions; Methods: Dataset and model evaluation; Methods: Dataset; Table 6
Profile review details
Targeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change.
Stable record: reported-task-0c92cda11228c4Specifications
Inputs, training, access and other details
Explanatory profile: limited source coverage · Automated source review, 2026-09-16. Review applies to the cited claims; unresolved fields are listed below. Numerical results retain their own review status.
| Property | Description and evidence |
|---|---|
| Datasets | SAbDab antibody–antigen complexes and the dMaSIF protein–protein interaction dataset.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
| Splits | The antibody–antigen collection is randomly partitioned with a validation subset and separate test complexes.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
| Metrics | ROC-AUC for the interaction task, comparing original and folded complex inputs; paratope/epitope metrics are separate evaluations.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
| Baselines | The interaction predictor is evaluated on original structures and complexes folded by AlphaFold3, Chai-1 and Boltz-1; these are alternative input-structure conditions.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
| Leakage controls | The source describes sequence deduplication and structural-similarity filtering before the random split.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
| Uncertainty | Table 5 reports folded-complex interaction AUC-ROC with ± terms, without defining their estimator in its caption. Tables 3–4 summarize folding RMSD and model confidence outputs separately; pLDDT, PAE and pTM are not confidence intervals on comparative interaction performance.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Tables 3–5 captions; Methods: Dataset and model evaluation |
| Entity type | Paper-specific computational evaluation protocol.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
| Organisms | The folding comparison uses antibody–antigen complexes selected from SAbDab. The dataset-construction text identifies structural and redundancy filters but does not provide a species census for the evaluated complex subset. Antibody and antigen origins must be distinguished. · Not reported in inspected sourcesSourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; Table 6 |
| Assays | Experimentally determined complex structures used for interaction assessment.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
| Allowed inputs | Folded antibody–antigen complexes.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
| Adaptation | Task training uses the curated complex training partition.SourcesEnhancing antibody-antigen interaction prediction with atomic flexibility · Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages |
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
Papers and result coverage
Last literature check: 2026-09-17. Primary-paper discovery and source inspection. Source-checked results are not independently reproduced experiments.
| Paper or primary resource | Version | Reference |
|---|---|---|
| Enhancing antibody-antigen interaction prediction with atomic flexibility | version of record | Read source DOI: 10.1371/journal.pcbi.1013576 |
Historical gaps recorded on 2026-09-17
The catalogue now holds 2 result rows for this benchmark. A note below about pending extraction describes the state on 2026-09-17 and may since have been answered by a later batch. The result rows and their sources are the current record.
- Original structure 0.89 ± 0.05 spans columns and is one reference result, not three independent results.
- Table 3 XML merges three RMSD cells using ampersands; numeric normalization needs explicit splitting/visual check.
- Confidence pTM/ipTM is not experimental interaction accuracy.
Search and extraction details
primary comparison table screened
Searches
- "antibody flexibility" "PMC12530544"
Evidence locations
- Tables 3–5 and captions
- Methods: folded-complex evaluation
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
17 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Diagram caption Conceptual summary of the cited evaluation; exact task configuration and source version remain part of the protocol. Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Diagram steps
| Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Diagram title Computational evaluation flow Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Datasets SAbDab antibody–antigen complexes and the dMaSIF protein–protein interaction dataset. Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Splits The antibody–antigen collection is randomly partitioned with a validation subset and separate test complexes. Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Adaptation Task training uses the curated complex training partition. Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Metrics ROC-AUC for the interaction task, comparing original and folded complex inputs; paratope/epitope metrics are separate evaluations. Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Baselines The interaction predictor is evaluated on original structures and complexes folded by AlphaFold3, Chai-1 and Boltz-1; these are alternative input-structure conditions. Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Leakage controls The source describes sequence deduplication and structural-similarity filtering before the random split. Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Uncertainty Table 5 reports folded-complex interaction AUC-ROC with ± terms, without defining their estimator in its caption. Tables 3–4 summarize folding RMSD and model confidence outputs separately; pLDDT, PAE and pTM are not confidence intervals on comparative interaction performance. Individual claims | Enhancing antibody-antigen interaction prediction with atomic flexibility Tables 3–5 captions; Methods: Dataset and model evaluation Version: version of record | source checked automated source review · 2026-09-16 Audit detailsTargeted full-paper and supplement review of the outstanding task fields, with original dataset metadata checked where accessible. Source-scoped omissions are explicit; no independent benchmark reproduction or numerical-result change. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Sources and history
View linked audit checks and correction history
Release 2026-09-29-06401fd5b220 · Record review: needs review
2 source records and release history
- Enhancing antibody-antigen interaction prediction with atomic flexibility · Original source · version of record
- Enhancing antibody-antigen interaction prediction with atomic flexibility · Original source · version of record
Technical metadata and extraction receipts
Stable ID: reported-task-0c92cda11228c4
- areas
- molecular-interactions
- tasks
- Antibody–antigen interaction prediction using folded complexes
- entity level
- task
- version
- Not reported
- task
- Antibody–antigen interaction prediction using folded complexes
- scope note
- Paper-specific evaluation task; protocol completeness requires further extraction.
- benchmark research
- review date: 2026-09-17; status: primary_comparison_table_screened; primary sources: expansion-p3-antibody-flexibility-2025; inspected locators: Tables 3–5 and captions; Methods: folded-complex evaluation; searched queries: "antibody flexibility" "PMC12530544"; gaps: Original structure 0.89 ± 0.05 spans columns and is one reference result, not three independent results.; Table 3 XML merges three RMSD cells using ampersands; numeric normalization needs explicit splitting/visual check.; Confidence pTM/ipTM is not experimental interaction accuracy.; claim scope: Primary-paper discovery and source inspection. Source-checked results are not independently reproduced experiments.
- historical missing metadata
- protocol version: not_reported_in_legacy_extract; split: not_reported_in_legacy_extract
- metadata review scope
- historical_missing_metadata preserves the original discovery state. Current descriptive evidence and missingness are recorded in profile.facts; numerical-result review is separate.
- legacy kinds
- benchmark
- entity classification
- review date: 2026-09-17; rationale: This source-scoped record identifies the biological prediction task and holds its paper context. Preserve the existing task identity; exact split, model adaptation and scoring remain in linked evaluations or separate protocol records.; source ids: antibody-flexibility-2025; source locator: Methods: Dataset; cached text lines 72–76; comparative evaluation and ablation passages; ambiguities: A paper- or suite-specific task may constrain some inputs or metrics; that alone does not make it interchangeable with a complete versioned protocol. No protocol equivalence is inferred.; Some legacy profile Entity type facts use the generic phrase computational evaluation protocol. That boilerplate is not sufficient to establish a single fixed protocol identity or to merge this task with another protocol record.