Predicted Bioactivity of Food Compounds Against BACE1 for Alzheimer's Disease

A dataset of 8,453 food-derived molecules screened for potential inhibitory activity against the BACE1 enzyme, a target for Alzheimer's disease therapy. The data was generated using a validated 6-parameter QSAR model and molecular docking studies, resulting in the identification of a primary hit compound. The dataset was created by Nobendu Mukerjee and last updated in May 2026.

Use Cases

• Virtual screening of natural compounds based on predicted pKi values.
• Training machine learning models for BACE1 inhibitor activity prediction using QSAR descriptors.
• Identifying structural motifs for drug design based on molecular docking results.
• Prioritizing food-derived molecules for experimental validation in Alzheimer's disease research.

Strengths

• Dataset is derived from an initial library of 26,467 food compounds, indicating a broad starting scope.
• The underlying QSAR model is reported to have acceptable statistical performance (R² = 0.82, Q²loo = 0.81).
• A specific hit molecule with a predicted pKi of 7.97 nM and binding affinity of -8.9 kcal/mol is identified.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the final downloadable data is unknown, which may limit suitability assessment.
• The dataset is very small (48.7 KB), suggesting limited scope or highly summarized results.

Provenance

Source

figshare

Collection Method

Generated via QSAR-based virtual screening and molecular docking of a food compound database.

Freshness

Last updated 2026-05-14 14:40:30; freshness should be verified.