Task

Protein–ligand binding affinity scoring

AK-score is evaluated as a protein–ligand scoring function using PDBbind and CASF tasks that separate scoring, ranking and pose selection.

Sources

AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63

37 evaluations · 82 results

Overview

Datasets

PDBbind-2016 refined complexes with affinity labels; the core set supplies the principal test complexes.

Metrics

Pearson correlation for scoring; Spearman, Kendall and predictive index for ranking; top-ranked pose success for docking.

Allowed inputs

Protein–ligand complex structures.

Sources

Evaluation 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

limited source coverage · Automated source review, 2026-09-16. All specifications and missing details

Results

Each comparison retains its reviewed evaluation scope, dataset and metric. Results are shown without a pooled ranking.

Evaluation scope and datasetMetric and reported setup

PDBbind-2016 core set · MAE

MAE (kcal/mol) · Lower values are better.

PDBbind-2016 core set (Protein–ligand binding affinity scoring) · PDBbind-2016 core set

Evidence origin: Independent external evaluation, Author-reported evaluation.

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

Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.

Comparison details and limitations

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.

Automated source review: 2026-09-17. Numerical source review does not establish independent reproduction.

Dots show point estimates. Whiskers show only explicitly defined uncertainty (standard deviation, standard error or a labelled interval); their definitions remain in Table. Unresolved uncertainty is not plotted. Differences do not establish statistical significance.

Find a tested model or method

Showing 9 of 9 matching rows.

Zoom to observed range

Tested configuration

11.061.121.171.23

Reported score

Methods and evaluation design

Procedure, tasks and evaluated configurations

How it works

Evaluation methodology

PDBbind-2016 refined complexes with affinity labels; the core set supplies the principal test complexes. The core set is excluded from the refined training set; an additional evaluation uses entries newly added in PDBbind-2018. Pearson correlation for scoring; Spearman, Kendall and predictive index for ranking; top-ranked pose success for docking. Reimplemented KDEEP, AutoDock Vina and X-score are evaluated in the paper. Exact core complexes are removed from training; protein-family or ligand-scaffold independence is not established by that exclusion. The paper describes bootstrap comparison of correlation coefficients.

Sources

Evaluation design

Benchmarks bring together tasks and protocols. A task describes the biological question; a protocol defines a particular test.

Protocols

These source-backed links do not make different protocols or scores interchangeable.

Recorded evaluations

Each evaluation records what was tested and under which conditions.

AK-score-single: Protein–ligand binding affinity scoring

Run this benchmark

Choose a concrete protocol before running an evaluation. Its inputs, split and scoring rules determine which results can be compared.

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

Core test complexes are removed from the affinity-training set, and CASF separates pose sampling from scoring.
Sources
AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63

Limitations and conditions

CASF scoring evaluates supplied structures and decouples scoring from pose sampling; it is not an end-to-end docking benchmark.
Sources
AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63

Profile review details

Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged.

Stable record: reported-task-a78312d5df6dad

Specifications

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.

Data, procedure and scoring
Property	Description and evidence
Datasets	PDBbind-2016 refined complexes with affinity labels; the core set supplies the principal test complexes. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Splits	The core set is excluded from the refined training set; an additional evaluation uses entries newly added in PDBbind-2018. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Metrics	Pearson correlation for scoring; Spearman, Kendall and predictive index for ranking; top-ranked pose success for docking. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Baselines	Reimplemented KDEEP, AutoDock Vina and X-score are evaluated in the paper. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Leakage controls	Exact core complexes are removed from training; protein-family or ligand-scaffold independence is not established by that exclusion. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Uncertainty	The paper describes bootstrap comparison of correlation coefficients. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Entity type	Paper-specific computational evaluation protocol. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Organisms	The benchmark selects PDBbind/CASF protein–ligand complexes by structural and affinity criteria. The dataset Methods and CASF evaluation section do not report a species-stratified inventory. · Not reported in inspected sources Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods: Protein–ligand data; CASF-2016 evaluation
Assays	Experimentally annotated protein–ligand affinities and structural poses. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Allowed inputs	Protein–ligand complex structures. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63
Adaptation	Supervised affinity scoring fitted on the refined set after excluding core test complexes. Sources AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63

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. Dated primary-source discovery and protocol/table screening. Source checking does not mean experimental reproduction. Only separately extracted and independently reviewed numeric batches are publishable.

Paper or primary resource	Version	Reference
AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks	version of record	Read source

Historical gaps recorded on 2026-09-17

The catalogue now holds 82 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.

exact checkpoint hashes and per-method scored denominators: Table labels alone do not establish these fields; do not infer checkpoint or scored count from model name or dataset size.

Search and extraction details

complete tables extracted

Searches

AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks 10.3390/ijms21228424

Evidence locations

Table 1; XML table ijms-21-08424-t001
Table 2; XML table ijms-21-08424-t002

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.

18 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-09-29-06401fd5b220
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	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.diagram.caption` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram steps Input: Protein–ligand complex structures. Evaluation: Supervised affinity scoring fitted on the refined set after excluding core test complexes. Readout: Pearson correlation for scoring; Spearman, Kendall and predictive index for ranking; top-ranked pose success for docking. Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.diagram.steps` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram title Computational evaluation flow Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.diagram.title` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Datasets PDBbind-2016 refined complexes with affinity labels; the core set supplies the principal test complexes. Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.0.value` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Splits The core set is excluded from the refined training set; an additional evaluation uses entries newly added in PDBbind-2018. Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.1.value` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Adaptation Supervised affinity scoring fitted on the refined set after excluding core test complexes. Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.10.value` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Metrics Pearson correlation for scoring; Spearman, Kendall and predictive index for ranking; top-ranked pose success for docking. Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.2.value` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Baselines Reimplemented KDEEP, AutoDock Vina and X-score are evaluated in the paper. Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.3.value` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Leakage controls Exact core complexes are removed from training; protein-family or ligand-scaffold independence is not established by that exclusion. Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.4.value` Source artifact SHA-256: `40cfd28dcd587599768ec99a6590ec593486475ff01c7b1d1f229b44aa91bf8d` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Uncertainty The paper describes bootstrap comparison of correlation coefficients. Individual claims	AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks Original source ↗ Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63 Version: version of record Retrieved: 2026-09-16T10:44:03.436853+00:00	source checked automated source review · 2026-09-16 Audit details Relevant full-paper computational evaluation sections, tables/captions and cited supplementary task passages were reviewed. Reporting omissions are scoped to the inspected sources. Original numerical results are unchanged. Field: `attributes.profile.facts.5.value` 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

View linked audit checks and correction history

Release 2026-09-29-06401fd5b220 · Record review: needs review

2 source records and release history

AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks · Original source · version of record
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: reported-task-a78312d5df6dad

areas: molecular-interactions
tasks: Protein–ligand binding affinity scoring
entity level: task
version: Not reported
task: Protein–ligand binding affinity scoring
scope note: Paper-specific evaluation task; protocol completeness requires further extraction.
comparison panels: id: akscore-2020-ijms-21-08424-t001-mae; title: PDBbind-2016 core set · MAE; protocol id: paper-protocol-d2055666ed2da8d7d7; dataset id: paper-dataset-b882219f61119f515b; metric: MAE; unit: kcal/mol; direction: lower; result ids: paper-result-2937b9a198c2ffe40c; paper-result-0cef59103b4c628552; paper-result-d80662bd72b1a95c2d; paper-result-1a479712bc41f21505; paper-result-554c59f6399b6ba950; paper-result-63d495bd9e454cfcd5; paper-result-4eab2cf242097371e4; paper-result-01a1e373e034e1d96b; paper-result-42921822c26167495c; source ids: akscore-2020; 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; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: akscore-2020-ijms-21-08424-t001-rmse; title: PDBbind-2016 core set · RMSE; protocol id: paper-protocol-d2055666ed2da8d7d7; dataset id: paper-dataset-b882219f61119f515b; metric: RMSE; unit: kcal/mol; direction: lower; result ids: paper-result-3c5790191bbab267ae; paper-result-23becb504fd247078e; paper-result-4b5b00557aa8abe03a; paper-result-d4f4f883f2596117e4; paper-result-bf54f8428bcdf902eb; paper-result-c924af293e1a3a8081; paper-result-8369e556ec65212435; paper-result-c9a3dfa12bfe1226e6; paper-result-7c4fb33077af93ea22; source ids: akscore-2020; 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 4: PDBbind-2016 core set RMSE; Table 1 (ijms-21-08424-t001), row 3 K DEEP, column 4: PDBbind-2016 core set RMSE; Table 1 (ijms-21-08424-t001), row 4 K DEEP, column 4: PDBbind-2016 core set RMSE; Table 1 (ijms-21-08424-t001), row 5 K DEEP, column 4: PDBbind-2016 core set RMSE; Table 1 (ijms-21-08424-t001), row 6 AK-score-single, column 4: PDBbind-2016 core set RMSE; Table 1 (ijms-21-08424-t001), row 7 AK-score-single, column 4: PDBbind-2016 core set RMSE; Table 1 (ijms-21-08424-t001), row 8 AK-score-single, column 4: PDBbind-2016 core set RMSE; Table 1 (ijms-21-08424-t001), row 9 AK-score-single, column 4: PDBbind-2016 core set RMSE; Table 1 (ijms-21-08424-t001), row 10 AK-score-ensemble, column 4: PDBbind-2016 core set RMSE; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: akscore-2020-ijms-21-08424-t002-pearson; title: CASF-2016 scoring · Pearson R; protocol id: paper-protocol-31d4bca49fb09d2e6f; dataset id: reported-dataset-f18fcc23dfa798; metric: Pearson R; unit: unitless; direction: higher; result ids: paper-result-9e734050de598079c9; paper-result-72483382e07392c9e8; paper-result-e2a9ae565f08fe24fe; paper-result-911b32fc8d0598d5be; paper-result-34a81b76f1d1a9be81; paper-result-133a5d56632431cd3b; paper-result-4858fe7af1b8734707; paper-result-3c0f9a1fd4492fa813; lit-b3-048; source ids: akscore-2020; source locator: A comparison of prediction accuracy with the CASF-2016 dataset.; Table 2 (ijms-21-08424-t002), row 3 K DEEP, column 3: CASF-2016 scoring Pearson R; Table 2 (ijms-21-08424-t002), row 4 K DEEP, column 3: CASF-2016 scoring Pearson R; Table 2 (ijms-21-08424-t002), row 5 K DEEP, column 3: CASF-2016 scoring Pearson R; Table 2 (ijms-21-08424-t002), row 6 K DEEP, column 3: CASF-2016 scoring Pearson R; Table 2 (ijms-21-08424-t002), row 7 AK-score-single, column 3: CASF-2016 scoring Pearson R; Table 2 (ijms-21-08424-t002), row 8 AK-score-single, column 3: CASF-2016 scoring Pearson R; Table 2 (ijms-21-08424-t002), row 9 AK-score-single, column 3: CASF-2016 scoring Pearson R; Table 2 (ijms-21-08424-t002), row 10 AK-score-single, column 3: CASF-2016 scoring Pearson R; Table 2 (ijms-21-08424-t002), row 11 AK-score-ensemble, column 3: CASF-2016 scoring Pearson R; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: akscore-2020-ijms-21-08424-t002-spearman; title: CASF-2016 ranking · Spearman correlation; protocol id: paper-protocol-0c5a1c5ef6896a2d96; dataset id: paper-dataset-1586129df956852448; metric: Spearman correlation; unit: unitless; direction: higher; result ids: paper-result-afde916d0d344fa5f3; paper-result-ed6bcd67134333f58a; paper-result-ebaae58e746eec83a7; paper-result-b2042668888da9601a; paper-result-ce979d714ded7834b1; paper-result-fcef1022a440b405a6; paper-result-85878ef679c27f7b5c; paper-result-e0b7d3b3917b02dad0; paper-result-7ea147a4415e34f1f0; source ids: akscore-2020; source locator: A comparison of prediction accuracy with the CASF-2016 dataset.; Table 2 (ijms-21-08424-t002), row 3 K DEEP, column 4: CASF-2016 ranking Spearman correlation; Table 2 (ijms-21-08424-t002), row 4 K DEEP, column 4: CASF-2016 ranking Spearman correlation; Table 2 (ijms-21-08424-t002), row 5 K DEEP, column 4: CASF-2016 ranking Spearman correlation; Table 2 (ijms-21-08424-t002), row 6 K DEEP, column 4: CASF-2016 ranking Spearman correlation; Table 2 (ijms-21-08424-t002), row 7 AK-score-single, column 4: CASF-2016 ranking Spearman correlation; Table 2 (ijms-21-08424-t002), row 8 AK-score-single, column 4: CASF-2016 ranking Spearman correlation; Table 2 (ijms-21-08424-t002), row 9 AK-score-single, column 4: CASF-2016 ranking Spearman correlation; Table 2 (ijms-21-08424-t002), row 10 AK-score-single, column 4: CASF-2016 ranking Spearman correlation; Table 2 (ijms-21-08424-t002), row 11 AK-score-ensemble, column 4: CASF-2016 ranking Spearman correlation; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: akscore-2020-ijms-21-08424-t002-kendall; title: CASF-2016 ranking · Kendall tau; protocol id: paper-protocol-0c5a1c5ef6896a2d96; dataset id: paper-dataset-1586129df956852448; metric: Kendall tau; unit: unitless; direction: higher; result ids: paper-result-7efbb2fd6cff6823e3; paper-result-6a2d74d436874496a7; paper-result-a1a3952f6bd9e2b283; paper-result-bb8ec19b9079636bfc; paper-result-c259bffd41e1648664; paper-result-5505231aa7e1d37837; paper-result-955fc84c22143025f1; paper-result-2eed37f01d919cd0e6; paper-result-38d90f85618bbf9bbb; source ids: akscore-2020; source locator: A comparison of prediction accuracy with the CASF-2016 dataset.; Table 2 (ijms-21-08424-t002), row 3 K DEEP, column 5: CASF-2016 ranking Kendall tau; Table 2 (ijms-21-08424-t002), row 4 K DEEP, column 5: CASF-2016 ranking Kendall tau; Table 2 (ijms-21-08424-t002), row 5 K DEEP, column 5: CASF-2016 ranking Kendall tau; Table 2 (ijms-21-08424-t002), row 6 K DEEP, column 5: CASF-2016 ranking Kendall tau; Table 2 (ijms-21-08424-t002), row 7 AK-score-single, column 5: CASF-2016 ranking Kendall tau; Table 2 (ijms-21-08424-t002), row 8 AK-score-single, column 5: CASF-2016 ranking Kendall tau; Table 2 (ijms-21-08424-t002), row 9 AK-score-single, column 5: CASF-2016 ranking Kendall tau; Table 2 (ijms-21-08424-t002), row 10 AK-score-single, column 5: CASF-2016 ranking Kendall tau; Table 2 (ijms-21-08424-t002), row 11 AK-score-ensemble, column 5: CASF-2016 ranking Kendall tau; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: akscore-2020-ijms-21-08424-t002-pi; title: CASF-2016 ranking · Predictive Index; protocol id: paper-protocol-0c5a1c5ef6896a2d96; dataset id: paper-dataset-1586129df956852448; metric: Predictive Index; unit: unitless; direction: higher; result ids: paper-result-5bc9b7cb53278a9249; paper-result-b1bed56b343f3c834f; paper-result-38ac6df2e198d58359; paper-result-5a60e66ec1c7bcaf4c; paper-result-e05720b0b94095d5af; paper-result-85c6bc19be77938986; paper-result-3d0c6443af15768910; paper-result-35759619206bf32ae8; paper-result-0d44e20dd3c089c7f3; source ids: akscore-2020; source locator: A comparison of prediction accuracy with the CASF-2016 dataset.; Table 2 (ijms-21-08424-t002), row 3 K DEEP, column 6: CASF-2016 ranking Predictive Index; Table 2 (ijms-21-08424-t002), row 4 K DEEP, column 6: CASF-2016 ranking Predictive Index; Table 2 (ijms-21-08424-t002), row 5 K DEEP, column 6: CASF-2016 ranking Predictive Index; Table 2 (ijms-21-08424-t002), row 6 K DEEP, column 6: CASF-2016 ranking Predictive Index; Table 2 (ijms-21-08424-t002), row 7 AK-score-single, column 6: CASF-2016 ranking Predictive Index; Table 2 (ijms-21-08424-t002), row 8 AK-score-single, column 6: CASF-2016 ranking Predictive Index; Table 2 (ijms-21-08424-t002), row 9 AK-score-single, column 6: CASF-2016 ranking Predictive Index; Table 2 (ijms-21-08424-t002), row 10 AK-score-single, column 6: CASF-2016 ranking Predictive Index; Table 2 (ijms-21-08424-t002), row 11 AK-score-ensemble, column 6: CASF-2016 ranking Predictive Index; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: akscore-2020-ijms-21-08424-t002-top1; title: CASF-2016 docking · Top 1 success; protocol id: paper-protocol-9e3344661a0f282e5e; dataset id: paper-dataset-30865ab989a9b4a55d; metric: Top 1 success; unit: percent; direction: higher; result ids: paper-result-3c9458f5230898c118; paper-result-b621ddd15cd16afefc; paper-result-c7f5dc9eb4c9ae58e6; paper-result-1aebc62b50d2a564cb; paper-result-045168c33da4f91423; paper-result-11080ec943cd58cc49; paper-result-2bb9d0ff2467197946; paper-result-76f65e01a22812dc8f; paper-result-9d866f174829993930; source ids: akscore-2020; source locator: A comparison of prediction accuracy with the CASF-2016 dataset.; Table 2 (ijms-21-08424-t002), row 3 K DEEP, column 7: CASF-2016 docking Top 1 success; Table 2 (ijms-21-08424-t002), row 4 K DEEP, column 7: CASF-2016 docking Top 1 success; Table 2 (ijms-21-08424-t002), row 5 K DEEP, column 7: CASF-2016 docking Top 1 success; Table 2 (ijms-21-08424-t002), row 6 K DEEP, column 7: CASF-2016 docking Top 1 success; Table 2 (ijms-21-08424-t002), row 7 AK-score-single, column 7: CASF-2016 docking Top 1 success; Table 2 (ijms-21-08424-t002), row 8 AK-score-single, column 7: CASF-2016 docking Top 1 success; Table 2 (ijms-21-08424-t002), row 9 AK-score-single, column 7: CASF-2016 docking Top 1 success; Table 2 (ijms-21-08424-t002), row 10 AK-score-single, column 7: CASF-2016 docking Top 1 success; Table 2 (ijms-21-08424-t002), row 11 AK-score-ensemble, column 7: CASF-2016 docking Top 1 success; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: akscore-2020-ijms-21-08424-t002-top2; title: CASF-2016 docking · Top 2 success; protocol id: paper-protocol-9e3344661a0f282e5e; dataset id: paper-dataset-30865ab989a9b4a55d; metric: Top 2 success; unit: percent; direction: higher; result ids: paper-result-696ea3d9b8dda19fd4; paper-result-e74ffc4f5d72ad3621; paper-result-6723f12f371db71f52; paper-result-928b3ac5d356652219; paper-result-ed9c485f143578778f; paper-result-97d06bfff5bd0deec7; paper-result-2257bde111aae5f357; paper-result-2aad15d58eda971fd3; paper-result-d88b7ac6133e0902be; source ids: akscore-2020; source locator: A comparison of prediction accuracy with the CASF-2016 dataset.; Table 2 (ijms-21-08424-t002), row 3 K DEEP, column 8: CASF-2016 docking Top 2 success; Table 2 (ijms-21-08424-t002), row 4 K DEEP, column 8: CASF-2016 docking Top 2 success; Table 2 (ijms-21-08424-t002), row 5 K DEEP, column 8: CASF-2016 docking Top 2 success; Table 2 (ijms-21-08424-t002), row 6 K DEEP, column 8: CASF-2016 docking Top 2 success; Table 2 (ijms-21-08424-t002), row 7 AK-score-single, column 8: CASF-2016 docking Top 2 success; Table 2 (ijms-21-08424-t002), row 8 AK-score-single, column 8: CASF-2016 docking Top 2 success; Table 2 (ijms-21-08424-t002), row 9 AK-score-single, column 8: CASF-2016 docking Top 2 success; Table 2 (ijms-21-08424-t002), row 10 AK-score-single, column 8: CASF-2016 docking Top 2 success; Table 2 (ijms-21-08424-t002), row 11 AK-score-ensemble, column 8: CASF-2016 docking Top 2 success; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17; id: akscore-2020-ijms-21-08424-t002-top3; title: CASF-2016 docking · Top 3 success; protocol id: paper-protocol-9e3344661a0f282e5e; dataset id: paper-dataset-30865ab989a9b4a55d; metric: Top 3 success; unit: percent; direction: higher; result ids: paper-result-f70d87df9407fc2486; paper-result-b4bea686f5bbd1459b; paper-result-5b72ab3d9513410a5f; paper-result-031247a72fb2604f0e; paper-result-7d7b1a31163a7d9b6b; paper-result-79f7e7e650d400e1bf; paper-result-1f1a756aa0ab1e0734; paper-result-e3010fc8650e08fdd7; paper-result-08702d09e3ca13e618; source ids: akscore-2020; source locator: A comparison of prediction accuracy with the CASF-2016 dataset.; Table 2 (ijms-21-08424-t002), row 3 K DEEP, column 9: CASF-2016 docking Top 3 success; Table 2 (ijms-21-08424-t002), row 4 K DEEP, column 9: CASF-2016 docking Top 3 success; Table 2 (ijms-21-08424-t002), row 5 K DEEP, column 9: CASF-2016 docking Top 3 success; Table 2 (ijms-21-08424-t002), row 6 K DEEP, column 9: CASF-2016 docking Top 3 success; Table 2 (ijms-21-08424-t002), row 7 AK-score-single, column 9: CASF-2016 docking Top 3 success; Table 2 (ijms-21-08424-t002), row 8 AK-score-single, column 9: CASF-2016 docking Top 3 success; Table 2 (ijms-21-08424-t002), row 9 AK-score-single, column 9: CASF-2016 docking Top 3 success; Table 2 (ijms-21-08424-t002), row 10 AK-score-single, column 9: CASF-2016 docking Top 3 success; Table 2 (ijms-21-08424-t002), row 11 AK-score-ensemble, column 9: CASF-2016 docking Top 3 success; context: 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.; caveats: Learning rates define distinct evaluated configurations. Table 2 ensemble Pearson 0.812 is distinct from the 30-network ensemble Pearson 0.827 in Figure 3. Do not transfer the 285-complex scoring denominator to ranking or docking rows without their appropriate grouping. Origins are labelled per method; appearance in one table does not constitute independent replication of every model.; review: method: automated_source_review; date: 2026-09-17
benchmark research: review date: 2026-09-17; status: complete_tables_extracted; primary sources: evidence-expansion-akscore-2020-40cfd28d; inspected locators: Table 1; XML table ijms-21-08424-t001; Table 2; XML table ijms-21-08424-t002; searched queries: AK-Score: Accurate Protein-Ligand Binding Affinity Prediction Using an Ensemble of 3D-Convolutional Neural Networks 10.3390/ijms21228424; gaps: exact checkpoint hashes and per-method scored denominators: Table labels alone do not establish these fields; do not infer checkpoint or scored count from model name or dataset size.; claim scope: Dated primary-source discovery and protocol/table screening. Source checking does not mean experimental reproduction. Only separately extracted and independently reviewed numeric batches are publishable.
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: akscore-2020; source locator: Methods §§3.1, 3.5; Results §2.3; Table 4; cached text lines 24, 27, 40–41, 59–63; 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.

Related records

dataset: CASF-2016
benchmark: AK-score-single: Protein–ligand binding affinity scoring
evaluates task: CASF-2016 ranking (Protein–ligand binding affinity scoring)
evaluates task: CASF-2016 scoring (Protein–ligand binding affinity scoring)
evaluates task: CASF-2016 docking (Protein–ligand binding affinity scoring)
evaluates task: PDBbind-2016 core set (Protein–ligand binding affinity scoring)

Suggest a correction