Evaluation

K DEEP · 0.0010: PDBbind-2016 core set

Training uses 3,772 PDBbind-2016 refined-set complexes after removal of the 285-complex core test set. CASF separates scoring, ranking and docking power.

Evaluation procedure

Training uses 3,772 PDBbind-2016 refined-set complexes after removal of the 285-complex core test set. CASF separates scoring, ranking and docking power.

Configuration: K DEEP · 0.0010
Protocol: PDBbind-2016 core set (Protein–ligand binding affinity scoring)
Dataset: PDBbind-2016 core set

origin: Independent external evaluation
configuration: K_DEEP reimplemented by AK-score authors in Keras 2.2.4 / TensorFlow 1.13.1; Table-specific network configuration; learning rate 0.0010
protocol id: paper-protocol-d2055666ed2da8d7d7
dataset version: Not reported
split: Training uses 3,772 PDBbind-2016 refined-set complexes after removal of the 285-complex core test set. CASF separates scoring, ranking and docking power.
subset: Not reported
population: count: 285; unit: protein–ligand complexes
aggregation: Not reported
inputs: Not reported
adaptation: K_DEEP reimplemented by AK-score authors in Keras 2.2.4 / TensorFlow 1.13.1; Table-specific network configuration; learning rate 0.0010
budget: Not reported
metric implementation: Not reported

Metadata review: needs review. Unreported conditions prevent automatic comparisons.

Evaluation results

Release 2026-09-17-d277315f7d76 · 1 evaluation · 2 metric rows. Different protocols are not a single leaderboard.

Evaluation setupMetricEvidence origin

Results grouped by the exact reported evaluation
Metric and finding	Coverage and uncertainty	Evidence
K DEEP · 0.0010: PDBbind-2016 core set Configuration: K DEEP · 0.0010Protocol: PDBbind-2016 core set (Protein–ligand binding affinity scoring)Dataset: PDBbind-2016 core set Training uses 3,772 PDBbind-2016 refined-set complexes after removal of the 285-complex core test set. CASF separates scoring, ranking and docking power. Independent external evaluation · Evaluation metadata: needs review
1.219 MAE Unit: kcal/mol · Direction: lower	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedAK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE Source checking is not independent reproduction.
1.536 RMSE Unit: kcal/mol · Direction: lower	Uncertainty: unreported Scored: Not reported · Eligible: Not reported	source checkedAK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 4: PDBbind-2016 core set RMSE Source checking is not independent reproduction.

Evidence table

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.

20 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-09-17-d277315f7d76
Property and statement	Original source and location	Review and provenance
attributes.comparison.adaptation K_DEEP reimplemented by AK-score authors in Keras 2.2.4 / TensorFlow 1.13.1; Table-specific network configuration; learning rate 0.0010 Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	not individually reviewed No individual claim review recorded independent paper Audit details Field: `attributes.comparison.adaptation` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.aggregation No value recorded Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	missing or unspecified No individual claim review recorded independent paper Audit details Field: `attributes.comparison.aggregation` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.budget No value recorded Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	missing or unspecified No individual claim review recorded independent paper Audit details Field: `attributes.comparison.budget` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.dataset_version No value recorded Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	missing or unspecified No individual claim review recorded independent paper Audit details Field: `attributes.comparison.dataset_version` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.inputs No value recorded Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	missing or unspecified No individual claim review recorded independent paper Audit details Field: `attributes.comparison.inputs` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.metric_implementation No value recorded Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	missing or unspecified No individual claim review recorded independent paper Audit details Field: `attributes.comparison.metric_implementation` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.population.count 285 Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	not individually reviewed No individual claim review recorded independent paper Audit details Field: `attributes.comparison.population.count` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.population.unit protein–ligand complexes Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	not individually reviewed No individual claim review recorded independent paper Audit details Field: `attributes.comparison.population.unit` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.protocol_id paper-protocol-d2055666ed2da8d7d7 Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	not individually reviewed No individual claim review recorded independent paper Audit details Field: `attributes.comparison.protocol_id` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
attributes.comparison.split Training uses 3,772 PDBbind-2016 refined-set complexes after removal of the 285-complex core test set. CASF separates scoring, ranking and docking power. Context-only references	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	not individually reviewed No individual claim review recorded independent paper Audit details Field: `attributes.comparison.split` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact

All evidence (CSV)All evidence (JSONL)How to read the evidence

Sources and history

Release 2026-09-17-d277315f7d76 · Record review: needs review

AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Original source · version of record

Download this release

Technical metadata and extraction receipts

Stable ID: paper-evaluation-e9df20dcf9d86b3201

areas: molecular-interactions
tasks: Protein–ligand binding affinity scoring
origin: independent_paper
protocol: Training uses 3,772 PDBbind-2016 refined-set complexes after removal of the 285-complex core test set. CASF separates scoring, ranking and docking power.
version: K_DEEP reimplemented by AK-score authors in Keras 2.2.4 / TensorFlow 1.13.1; Table-specific network configuration; learning rate 0.0010
comparison: protocol id: paper-protocol-d2055666ed2da8d7d7; dataset version: Not reported; split: Training uses 3,772 PDBbind-2016 refined-set complexes after removal of the 285-complex core test set. CASF separates scoring, ranking and docking power.; subset: Not reported; population: count: 285; unit: protein–ligand complexes; aggregation: Not reported; inputs: Not reported; adaptation: K_DEEP reimplemented by AK-score authors in Keras 2.2.4 / TensorFlow 1.13.1; Table-specific network configuration; learning rate 0.0010; budget: Not reported; metric implementation: Not reported
source locator: Assessment of prediction accuracy of ResNext-ensemble, ResNext, and K DEEP using the PDBbind-2016 dataset and mean absolute error and root mean square error metrics.; Table 1 (ijms-21-08424-t001), row 2 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 3: PDBbind-2016 core set MAE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 3: PDBbind-2016 core set MAE
missing metadata: checkpoint revision: unextracted; budget: unreported; split manifest: unextracted

Related records

model: K DEEP · 0.0010
benchmark: PDBbind-2016 core set (Protein–ligand binding affinity scoring)
dataset: PDBbind-2016 core set
evaluation: K DEEP: MAE on PDBbind-2016 core set
evaluation: K DEEP: RMSE on PDBbind-2016 core set

Suggest a correction