Computational Antibody-Antigen Binding Predictions for 14 Protein Complexes

A computational dataset supporting a strategy to predict antibody-antigen binding, featuring 14 different antibody-bound complexes with 9 diverse antigens. The data was generated by Mehmet Emin Aygen and includes results from essential site scanning analysis (ESSA), molecular docking with ClusPro, and molecular dynamics simulations. The dataset was last updated on 2026-05-27.

Use Cases

• Benchmarking epitope prediction methods based on the ESSA-guided docking results for Interleukin-1β.
• Training machine learning models for binding pose prediction using the molecular dynamics simulation results.
• Validating coarse-grained modeling efficiency for rapid screening of key binding regions as described in the workflow.

Strengths

• Includes a rigorous test case with Interleukin-1β (IL-1β) featuring four distinct epitope regions.
• Compares results with AlphaFold3 predictions, providing a modern benchmark.
• Dataset size is 174.9 MB, suggesting substantial computational output.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare

Collection Method

Generated via computational strategy including ESSA, molecular docking, and molecular dynamics simulations.

Freshness

Last updated 2026-05-27 15:06:44; freshness should be verified.