dataset · audited-with-caveats · verified 2026-08-14

AB-Bind

Antibody-focused mutational binding data with accompanying structures for computational affinity prediction.

Audited with caveats: 2 field(s) are marked provisional or conflicted. Warnings are shown next to affected values and these claims are excluded from unqualified comparisons.

Benchmark definition

What is counted

Version: initial-release
Total: 1101 (experimentally determined binding free-energy changes across the complete database)
Task formats: Predict mutation-induced binding free-energy changes; Classify mutants as improved or weakened binders; Rank candidate mutations for improved affinity
Capabilities: PredictionClassificationRegressionOptimization
Modalities: Table3D structureWet-lab output

Version history

Version	Status	Release / as-of	Total	Formal tracks
initial-release `ab-bind-initial-release-version`	current	2015-11-06	1101 (experimentally determined binding free-energy changes across the complete database)	None registered

Scientific Task Atlas

Scientific task classification

partial for initial-release. No Scientific Task claim passed independent high-confidence verification; task mapping remains pending a targeted official-source audit.

The source names only a broad direction; no more specific leaf task can be assigned without inference.

Relationship registry

How works use this benchmark

Partial claims, non-evaluation uses, and third-party summaries stay visible without entering model comparisons.

Partial evaluation claims

evaluation

ab-bind-antibody-binding-mutational-database-for-compu-ab-bind-2-use

Work: AB-Bind: Antibody binding mutational database for computational affinity predictions · source version ab-bind-antibody-binding-mutational-database-for-compu-2015-11-06

Selection: not reported
Models: Discovery Studio, dDFIRE, DFIRE, FoldX, Rosetta, STATIUM, bASA
Metrics: Not reported / not applicable
Linked runs: None

Not reported / unresolved: No random seed is reported.; Realized n is not printed for the low- and medium-confidence ROC subsets.; Several method-specific Pearson correlations and confidence intervals appear only in referenced Supporting Information.; Provider identities are not printed for several scoring methods.; benchmark version; realized n/scope; metric; numeric result; prompt and tools; grader and repeats

Owner-reviewed conservative publication: the creator evaluation is retained only as a partial relationship; conflicted settings and outcomes are omitted pending manual reconciliation.

Evidence

section: Results > Scoring potentials
Supports: /relation_type
section: Discussion
Supports: /benchmark_id
section: Results > Scoring potentials
Supports: /model_ids
section: Results > Scoring potentials
Supports: /model_ids
section: Results > Scoring potentials
Supports: /model_ids
section: Results > Scoring potentials
Supports: /model_ids
section: Materials and Methods > Structure generation
Supports: /model_ids
section: Materials and Methods > Structure generation
Supports: /model_ids
other: Document pages 13 and 17, Structure generation and Reference 61
Supports: /model_ids

evaluation

explainable-protein-protein-binding-affinity-predictio-ab-bind-5-use

Work: Explainable protein-protein binding affinity prediction via fine-tuning protein language models · source version explainable-protein-protein-binding-affinity-predictio-biorxiv-version-posted-2

Selection: filtered · Removed parent-PDB overlap or identical antigen-antibody sequence-pair overlap with PPB-Affinity training data
Models: BALM-PPI
Metrics: Not reported / not applicable
Linked runs: None

Not reported / unresolved: Benchmark version is not reported.; BALM-PPI provider and version are not explicitly reported.; benchmark version; metric; numeric result

AI-assisted double-pass extraction; values are limited to independently supported claims.

Evidence

section: 2.3
Supports: /relation_type
section: 2.3
Supports: /benchmark_id
section: 2.3
Supports: /scope
section: 2.3
Supports: /scope
table: Table 1
Supports: /scope
section: 2.3
Supports: /scope
table: Table 1
Supports: /model_ids

Creation, training, validation, or model-selection uses

benchmark creation

ab-bind-antibody-binding-mutational-database-for-compu-ab-bind-1-use

Work: AB-Bind: Antibody binding mutational database for computational affinity predictions · source version ab-bind-antibody-binding-mutational-database-for-compu-2015-11-06

Selection: not applicable
Models: Not reported / not applicable
Metrics: Not reported / not applicable
Linked runs: None

Not reported / unresolved: No explicit AB-Bind version identifier is reported.; The official repository metadata reports no license.; The source does not reconcile the conflicting broad multiple-mutation count with its detailed partition.

AI-assisted double-pass extraction; values are limited to independently supported claims.

Evidence

section: Results > AB-Bind database
Supports: /relation_type
section: Abstract
Supports: /benchmark_id

fine tuning

ab-bind-antibody-binding-mutational-database-for-compu-ab-bind-3-use

Work: AB-Bind: Antibody binding mutational database for computational affinity predictions · source version ab-bind-antibody-binding-mutational-database-for-compu-2015-11-06

Selection: not applicable
Models: FoldX
Metrics: Not reported / not applicable
Linked runs: None

Not reported / unresolved: The fitting split, realized n, regression specification, and numeric results are not reported; the paper says data not shown.

AI-assisted double-pass extraction; values are limited to independently supported claims.

Evidence

section: Discussion
Supports: /relation_type
section: Discussion
Supports: /benchmark_id
other: Document pages 12–13, Discussion and Structure generation
Supports: /model_ids
section: Discussion
Supports: /scope

fine tuning

explainable-protein-protein-binding-affinity-predictio-ab-bind-6-use

Selection: not applicable · Few-shot 10%
Models: BALM-PPI
Metrics: Not reported / not applicable
Linked runs: None

Not reported / unresolved: Benchmark version and exact realized training/evaluation n are not reported.; BALM-PPI provider and version are not explicitly reported.

AI-assisted double-pass extraction; values are limited to independently supported claims.

Evidence

table: Table 1
Supports: /relation_type
table: Table 1
Supports: /benchmark_id
table: Table 1
Supports: /scope
table: Table 1
Supports: /scope
table: Table 1
Supports: /scope
table: Table 1
Supports: /model_ids

fine tuning

explainable-protein-protein-binding-affinity-predictio-ab-bind-7-use

Selection: not applicable · Few-shot 20%
Models: BALM-PPI
Metrics: Not reported / not applicable
Linked runs: None

Not reported / unresolved: Benchmark version and exact realized training/evaluation n are not reported.; BALM-PPI provider and version are not explicitly reported.

AI-assisted double-pass extraction; values are limited to independently supported claims.

Evidence

table: Table 1
Supports: /relation_type
table: Table 1
Supports: /benchmark_id
table: Table 1
Supports: /scope
table: Table 1
Supports: /scope
table: Table 1
Supports: /scope
table: Table 1
Supports: /model_ids

fine tuning

explainable-protein-protein-binding-affinity-predictio-ab-bind-8-use

Selection: not applicable · Few-shot 30%
Models: BALM-PPI
Metrics: Not reported / not applicable
Linked runs: None

Not reported / unresolved: Benchmark version and exact realized training/evaluation n are not reported.; BALM-PPI provider and version are not explicitly reported.

AI-assisted double-pass extraction; values are limited to independently supported claims.

Evidence

table: Table 1
Supports: /relation_type
table: Table 1
Supports: /benchmark_id
table: Table 1
Supports: /scope
table: Table 1
Supports: /scope
table: Table 1
Supports: /scope
table: Table 1
Supports: /model_ids

Evaluation registry

Works and run settings

A setting change—scope, prompt, tools, budget, grader, or repeats—creates a separate run. Charts never cross a comparability group.

No normalized evaluation run is published yet. Creator evidence is still attached below.

Evidence and change history

Source locators remain visible; expand an item to inspect the exact Registry fields it supports.

AB-Bind: Antibody binding mutational database for computational affinity predictions · repository-path: sarahsirin/AB-Bind-Database@f5af13df80000ad9e438a6ccf1147bd5f9d55dba > file_paths · Supports 1 field

What is counted

Version history

Scientific task classification

How works use this benchmark

Partial evaluation claims

ab-bind-antibody-binding-mutational-database-for-compu-ab-bind-2-use

explainable-protein-protein-binding-affinity-predictio-ab-bind-5-use

Creation, training, validation, or model-selection uses

ab-bind-antibody-binding-mutational-database-for-compu-ab-bind-1-use

ab-bind-antibody-binding-mutational-database-for-compu-ab-bind-3-use

explainable-protein-protein-binding-affinity-predictio-ab-bind-6-use

explainable-protein-protein-binding-affinity-predictio-ab-bind-7-use

explainable-protein-protein-binding-affinity-predictio-ab-bind-8-use

Works and run settings

Evidence and change history

Unresolved field claims