Model type
Trainable AlphaFold2-compatible structure predictor
Configuration
OpenFold
OpenFold is a trainable PyTorch implementation of AlphaFold 2 and AlphaFold-Multimer workflows.
Sources (2)
aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright NoticeHow it worksOpenFold workflow
Conceptual summary of the documented data flow; optional inputs and configured downstream stages must be reported for a reproducible evaluation.
Sources (2)
aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright NoticeOverview
Inputs
Protein sequence, sequence alignments and optional structural templates, depending on the inference mode.
Outputs
Predicted protein structures from selected OpenFold or compatible AlphaFold parameters.
Access
Official project documentation and implementation: https://github.com/aqlaboratory/openfold
Sources (2)
aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Noticelimited source coverage · Automated source review, 2026-09-16. All specifications and missing details
Evaluations and results
14 evaluations · 14 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: OpenFold | Task: ProteinBench FOLD-ACCURACY-GDT-TS-MEAN: Accuracy GDT-TS (mean) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0.856 gdt-ts_mean score · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-ACCURACY-GDT-TS-MEAN: Accuracy GDT-TS (mean) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Accuracy GDT-TS ↑) |
| Configuration: OpenFold | Task: ProteinBench FOLD-ACCURACY-GDT-TS-MEDIAN: Accuracy GDT-TS (median) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0.913 gdt-ts_median score · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-ACCURACY-GDT-TS-MEDIAN: Accuracy GDT-TS (median) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Accuracy GDT-TS ↑) |
| Configuration: OpenFold | Task: ProteinBench FOLD-ACCURACY-LDDT-MEAN: Accuracy lDDT (mean) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0.899 lddt_mean score · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-ACCURACY-LDDT-MEAN: Accuracy lDDT (mean) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Accuracy lDDT ↑) |
| Configuration: OpenFold | Task: ProteinBench FOLD-ACCURACY-LDDT-MEDIAN: Accuracy lDDT (median) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0.933 lddt_median score · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-ACCURACY-LDDT-MEDIAN: Accuracy lDDT (median) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Accuracy lDDT ↑) |
| Configuration: OpenFold | Task: ProteinBench FOLD-ACCURACY-RMSD-MEAN: Accuracy RMSD (mean) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 3.21 rmsd_mean score · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-ACCURACY-RMSD-MEAN: Accuracy RMSD (mean) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Accuracy RMSD ↓) |
| Configuration: OpenFold | Task: ProteinBench FOLD-ACCURACY-RMSD-MEDIAN: Accuracy RMSD (median) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 1.59 rmsd_median score · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-ACCURACY-RMSD-MEDIAN: Accuracy RMSD (median) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Accuracy RMSD ↓) |
| Configuration: OpenFold | Task: ProteinBench FOLD-ACCURACY-TM-SCORE-MEAN: Accuracy TM-score (mean) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0.87 tm-score_mean score · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-ACCURACY-TM-SCORE-MEAN: Accuracy TM-score (mean) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Accuracy TM-score ↑) |
| Configuration: OpenFold | Task: ProteinBench FOLD-ACCURACY-TM-SCORE-MEDIAN: Accuracy TM-score (median) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0.947 tm-score_median score · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-ACCURACY-TM-SCORE-MEDIAN: Accuracy TM-score (median) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Accuracy TM-score ↑) |
| Configuration: OpenFold | Task: ProteinBench FOLD-QUALITY-CA-BREAK-MEAN: Quality CA break (%) (mean) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0% ca-break_mean percent · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-QUALITY-CA-BREAK-MEAN: Quality CA break (%) (mean) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Quality CA break (%) ↓) |
| Configuration: OpenFold | Task: ProteinBench FOLD-QUALITY-CA-BREAK-MEDIAN: Quality CA break (%) (median) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0% ca-break_median percent · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-QUALITY-CA-BREAK-MEDIAN: Quality CA break (%) (median) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Quality CA break (%) ↓) |
| Configuration: OpenFold | Task: ProteinBench FOLD-QUALITY-CA-CLASH-MEAN: Quality CA clash (%) (mean) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0.4% ca-clash_mean percent · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-QUALITY-CA-CLASH-MEAN: Quality CA clash (%) (mean) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Quality CA clash (%) ↓) |
| Configuration: OpenFold | Task: ProteinBench FOLD-QUALITY-CA-CLASH-MEDIAN: Quality CA clash (%) (median) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 0% ca-clash_median percent · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-QUALITY-CA-CLASH-MEDIAN: Quality CA clash (%) (median) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Quality CA clash (%) ↓) |
| Configuration: OpenFold | Task: ProteinBench FOLD-QUALITY-PEPBOND-BREAK-MEAN: Quality PepBond break (%) (mean) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 2% pepbond-break_mean percent · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-QUALITY-PEPBOND-BREAK-MEAN: Quality PepBond break (%) (mean) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Quality PepBond break (%) ↓) |
| Configuration: OpenFold | Task: ProteinBench FOLD-QUALITY-PEPBOND-BREAK-MEDIAN: Quality PepBond break (%) (median) Dataset subset: CAMEO2022, 183 proteins (ProteinBench split) | 1.7% pepbond-break_median percent · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourceOpenFold on ProteinBench FOLD-QUALITY-PEPBOND-BREAK-MEDIAN: Quality PepBond break (%) (median) Structure prediction on CAMEO2022. Each cell prints the mean and the median over 183 proteins. Aggregation: Not reported proteinbench primary benchmark evidence · Table 7, row(OpenFold), column(Quality PepBond break (%) ↓) |
Source checking is not independent reproduction. Release 2026-09-29-06401fd5b220.
Use this model
How it works, versions and access
Related profile: OpenFold. This page retains the exact record and its evaluation context.
This configuration
Protein model evaluated by the ProteinBench authors under their harness.
- record
- OpenFold
- configuration
- Not reported
- entity type
- Configuration
How it works
How it works
OpenFold is a trainable PyTorch implementation of AlphaFold 2 and AlphaFold-Multimer workflows. AlphaFold-compatible folding architecture with configurable attention implementations and a training pipeline. The documented inputs are protein sequence, sequence alignments and optional structural templates, depending on the inference mode. The output consists of predicted protein structures from selected OpenFold or compatible AlphaFold parameters.
Sources (2)
aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright NoticeVersions and reproducibility
Documented monomer v2.0.1 and multimer v2.3.2 compatibility; distinguish OpenFold and imported AlphaFold parameters. Memory- and configuration-dependent; low-memory attention and CPU offloading are documented.
Sources (2)
aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright NoticeStrengths, limitations and unresolved questions
Strengths and limitations
Strengths and considerations
- Provides training code, alignment resources and conversion between compatible parameter formats.
Sources (2)
aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice
Limitations and conditions
- OpenFold-trained weights and DeepMind weights are separate artifacts. The documentation notes deliberate differences, including omitted model ensembling.
Sources (2)
aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice
Profile review details
Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied.
Stable record: discovery-model-openfoldSpecifications
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 |
|---|---|
| Model type | Trainable AlphaFold2-compatible structure predictorSources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Architecture | AlphaFold-compatible folding architecture with configurable attention implementations and a training pipeline.Sources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Inputs | Protein sequence, sequence alignments and optional structural templates, depending on the inference mode.Sources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Outputs | Predicted protein structures from selected OpenFold or compatible AlphaFold parameters.Sources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Parameters | The official documentation supports different AlphaFold2-compatible configurations and heads; this family record does not select one checkpoint with one verified total. · Not reported in inspected sourcesSources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Known versions | Documented monomer v2.0.1 and multimer v2.3.2 compatibility; distinguish OpenFold and imported AlphaFold parameters.Sources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Training data | The documentation releases approximately 400,000 MSAs and PDB70 template-hit files; the actual training schedule/checkpoint remains a separate identity.Sources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Training cutoff | The inspected implementation documentation does not supply one common cutoff across its supported original, retrained and extended-context model configurations. · Not reported in inspected sourcesSources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Context limits | Memory- and configuration-dependent; low-memory attention and CPU offloading are documented.Sources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Weights licence | The copyright notice specifies CC-BY-4.0 for DeepMind pretrained parameters; this does not automatically establish all OpenFold checkpoint terms.Sources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Access | Official project documentation and implementation: https://github.com/aqlaboratory/openfoldSources (2)aqlaboratory/openfold: README.md; aqlaboratory/openfold: docs/source/original_readme.md · docs/source/original_readme.md: Features, Inference and Copyright Notice |
| Code licence | Apache-2.0Sourcesaqlaboratory/openfold: LICENSE · LICENSE: licence text |
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
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.
1 evidence row matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Relationship: family discovery-model-openfold Individual claims | proteinbench primary benchmark evidence Tables 7; task methods and corresponding named row Version: 2409.06744v1 | source checked automated source review · 2026-09-23 Audit detailsSource review establishes this relationship only. Exact evaluated configurations and original numerical review status remain unchanged. Paper evaluates this named model under its ProteinBench harness. Association is to the family, not an assertion of checkpoint equivalence or cross-task comparability. Field: Claim: model-evaluation-identity-d9ecb01e63465c934c3f Source artifact SHA-256: Hash scope: Exact retrieved primary paper artifact bytes. |
Sources and history
View linked audit checks and correction history
Release 2026-09-29-06401fd5b220 · Record review: source checked
1 source records and release history
- proteinbench primary benchmark evidence · Original source · 2409.06744v1
Technical metadata and extraction receipts
Stable ID: proteinbench-method-openfold
- areas
- proteins-complexes
- source locator
- Table 7, row(OpenFold)
- missing metadata
- checkpoint revision: unreported; parameters: unextracted
Related records
- family: OpenFold
- subject: OpenFold: family OpenFold
- model: OpenFold on ProteinBench FOLD-ACCURACY-GDT-TS-MEAN: Accuracy GDT-TS (mean)
- model: OpenFold on ProteinBench FOLD-ACCURACY-GDT-TS-MEDIAN: Accuracy GDT-TS (median)
- model: OpenFold on ProteinBench FOLD-ACCURACY-LDDT-MEAN: Accuracy lDDT (mean)
- model: OpenFold on ProteinBench FOLD-ACCURACY-LDDT-MEDIAN: Accuracy lDDT (median)
- model: OpenFold on ProteinBench FOLD-ACCURACY-RMSD-MEAN: Accuracy RMSD (mean)
- model: OpenFold on ProteinBench FOLD-ACCURACY-RMSD-MEDIAN: Accuracy RMSD (median)
- model: OpenFold on ProteinBench FOLD-ACCURACY-TM-SCORE-MEAN: Accuracy TM-score (mean)
- model: OpenFold on ProteinBench FOLD-ACCURACY-TM-SCORE-MEDIAN: Accuracy TM-score (median)
- model: OpenFold on ProteinBench FOLD-QUALITY-CA-BREAK-MEAN: Quality CA break (%) (mean)
- model: OpenFold on ProteinBench FOLD-QUALITY-CA-BREAK-MEDIAN: Quality CA break (%) (median)
- model: OpenFold on ProteinBench FOLD-QUALITY-CA-CLASH-MEAN: Quality CA clash (%) (mean)
- model: OpenFold on ProteinBench FOLD-QUALITY-CA-CLASH-MEDIAN: Quality CA clash (%) (median)
- model: OpenFold on ProteinBench FOLD-QUALITY-PEPBOND-BREAK-MEAN: Quality PepBond break (%) (mean)
- model: OpenFold on ProteinBench FOLD-QUALITY-PEPBOND-BREAK-MEDIAN: Quality PepBond break (%) (median)