Protocol

PLABench KIBA, cold-target split

Affinity regression on KIBA, cold-target split: test proteins unseen in training and validation; models retrained by the benchmark authors.

3 evaluations · 18 results

Overview

An explanatory profile has not yet been reviewed for this record.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

3 recorded evaluations, 18 metric rows. A comparison chart has not yet been validated for these results. The table retains the individual findings and their sources.

View coverage and remaining gaps across all benchmarks

Results

Results are available, but no reviewed comparison panel is linked in this release.

All evaluations

3 evaluations · 18 results. Different protocols are not a single leaderboard.

Protocol: PLABench KIBA, cold-target split · Dataset: KIBA cold-target test set (PLABench)

Sorted by Concordance index (PLABench KIBA-cold-target) (higher is better). The best value in each column is highlighted. Decimals are rounded for display; each value links to the printed value and its source.

Tested configuration	Concordance index (PLABench KIBA-cold-target)	Modified squared correlation (PLABench KIBA-cold-target)	MSE (PLABench KIBA-cold-target)	Pearson (PLABench KIBA-cold-target)	RMSE (PLABench KIBA-cold-target)	Spearman (PLABench KIBA-cold-target)
LLF (retrained on Davis and KIBA, PLABench)Evaluation	0.795 (best in column)	0.523 (best in column)	0.343 (best in column)	0.726 (best in column)	0.585 (best in column)	0.735 (best in column)
MixingDTA (retrained on Davis and KIBA, PLABench)Evaluation	0.761	0.502	0.358	0.712	0.598	0.656
DeepDTA (retrained on Davis and KIBA, PLABench)Evaluation	0.698	0.297	0.511	0.558	0.714	0.522

All 18 result rows with coverage, uncertainty and sources

Filter evaluations

Tested model or methodProtocolDatasetEvaluation setupMetricEvidence origin

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configuration	Protocol and dataset	Finding	Evidence and details
Configuration: DeepDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.698 Concordance index (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'DeepDTA', strategy 'Cold-target', dataset 'KIBA', column 'CI' (text-layer line 442)
Configuration: DeepDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.511 MSE (PLABench KIBA-cold-target) unit-unreported · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'DeepDTA', strategy 'Cold-target', dataset 'KIBA', column 'MSE' (text-layer line 442)
Configuration: DeepDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.558 Pearson (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'DeepDTA', strategy 'Cold-target', dataset 'KIBA', column 'Pearson' (text-layer line 442)
Configuration: DeepDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.297 Modified squared correlation (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'DeepDTA', strategy 'Cold-target', dataset 'KIBA', column 'rm2' (text-layer line 442)
Configuration: DeepDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.714 RMSE (PLABench KIBA-cold-target) unit-unreported · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'DeepDTA', strategy 'Cold-target', dataset 'KIBA', column 'RMSE' (text-layer line 442)
Configuration: DeepDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.522 Spearman (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'DeepDTA', strategy 'Cold-target', dataset 'KIBA', column 'Spearman' (text-layer line 442)
Configuration: LLF (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.795 Concordance index (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source LLF on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'LLF', strategy 'Cold-target', dataset 'KIBA', column 'CI' (text-layer line 443)
Configuration: LLF (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.343 MSE (PLABench KIBA-cold-target) unit-unreported · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source LLF on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'LLF', strategy 'Cold-target', dataset 'KIBA', column 'MSE' (text-layer line 443)
Configuration: LLF (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.726 Pearson (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source LLF on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'LLF', strategy 'Cold-target', dataset 'KIBA', column 'Pearson' (text-layer line 443)
Configuration: LLF (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.523 Modified squared correlation (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source LLF on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'LLF', strategy 'Cold-target', dataset 'KIBA', column 'rm2' (text-layer line 443)
Configuration: LLF (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.585 RMSE (PLABench KIBA-cold-target) unit-unreported · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source LLF on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'LLF', strategy 'Cold-target', dataset 'KIBA', column 'RMSE' (text-layer line 443)
Configuration: LLF (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.735 Spearman (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source LLF on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'LLF', strategy 'Cold-target', dataset 'KIBA', column 'Spearman' (text-layer line 443)
Configuration: MixingDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.761 Concordance index (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source MixingDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'MixingDTA', strategy 'Cold-target', dataset 'KIBA', column 'CI' (text-layer line 444)
Configuration: MixingDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.358 MSE (PLABench KIBA-cold-target) unit-unreported · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source MixingDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'MixingDTA', strategy 'Cold-target', dataset 'KIBA', column 'MSE' (text-layer line 444)
Configuration: MixingDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.712 Pearson (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source MixingDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'MixingDTA', strategy 'Cold-target', dataset 'KIBA', column 'Pearson' (text-layer line 444)
Configuration: MixingDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.502 Modified squared correlation (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source MixingDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'MixingDTA', strategy 'Cold-target', dataset 'KIBA', column 'rm2' (text-layer line 444)
Configuration: MixingDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.598 RMSE (PLABench KIBA-cold-target) unit-unreported · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source MixingDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'MixingDTA', strategy 'Cold-target', dataset 'KIBA', column 'RMSE' (text-layer line 444)
Configuration: MixingDTA (retrained on Davis and KIBA, PLABench)	Protocol: PLABench KIBA, cold-target split Dataset: KIBA cold-target test set (PLABench)	0.656 Spearman (PLABench KIBA-cold-target) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source MixingDTA on PLABench KIBA-cold-target lit-interactions-20261011-leakctl-protocol-kiba-cold-target Aggregation: Not reported Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Table 2, row 'MixingDTA', strategy 'Cold-target', dataset 'KIBA', column 'Spearman' (text-layer line 444)

Source checking is not independent reproduction. Release 2026-10-23-5347695500ed.

Methods and evaluation design

Procedure, tasks and evaluated configurations

Recorded evaluations

Each evaluation records what was tested and under which conditions.

Baseline coverage

Reference methods help show what a model adds beyond simple controls. We track a null control and a conventional method for each protocol.

0 of 2 active baseline roles have published Rewire measurements in this release. Measurements on a selected protocol do not establish coverage of an entire suite.

No execution recipe linked to this protocol. Recipe availability does not establish a completed evaluation.

External evaluations: 3

Literature evidence is not a Rewire measurement. Executed but unpublished runs and private review status are not included.

Null control

Proposed control: requires review

Training-set mean where supervised fitting is permitted

Protocol-specific applicability, permitted inputs, access, split, evaluator and execution requirements need review before implementation or execution.

This is a suggested selection rule, not a validated method or a measured score.

Conventional reference

Proposed control: requires review

Simple features with train-only ridge regression

Protocol-specific applicability, permitted inputs, access, split, evaluator and execution requirements need review before implementation or execution.

This is a suggested selection rule, not a validated method or a measured score.

Protocol coverage CSV (gzip) · Model evaluation matrix (gzip) · Source table (gzip) · Release and checksums (gzip)

Coverage is derived from release 2026-10-23-5347695500ed. Source citations describe the original records; they do not validate an unreviewed baseline proposal. No results have been generated by this audit.

Run instructions

No runnable recipe has been reviewed for this protocol. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.

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.

0 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-10-23-5347695500ed
Property and statement	Original source and location	Review and provenance

No evidence rows match these filters. Choose another scope or clear the search.

How to read the evidence

Sources and history

Release 2026-10-23-5347695500ed · Record review: source checked

1 source record and release history

Leakage-controlled benchmarking reveals generalization limits of deep learning for protein-ligand binding affinity prediction · Original source · bioRxiv 10.64898/2026.09.17.752433v1 (posted 2026-09-23); 30-page PDF

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: lit-interactions-20261011-leakctl-protocol-kiba-cold-target

areas: molecular-interactions
contexts: research
protocol: KIBA cold-target test set; MSE, RMSE, Pearson, Spearman, rm2 and CI as printed in Table 2
version: PLABench bioRxiv v1
limitations: Metrics are computed over all test pairs of the set (Methods, 'Evaluation metrics'); per-target averaging is not stated for these tables.; Only one value per cell is printed; no uncertainty is reported in Tables 2-4.; Only the three sequence-based models (DeepDTA, LLF, MixingDTA) were retrained on Davis and KIBA.
source locator: Table 2, 'KIBA' columns, strategy 'Cold-target'

Related records

part of: PLABench
uses data: KIBA cold-target test set (PLABench)
assessment: DeepDTA on PLABench KIBA-cold-target
assessment: LLF on PLABench KIBA-cold-target
assessment: MixingDTA on PLABench KIBA-cold-target

Suggest a correction