Dietary Flavonoid Network Pharmacology Predictions with Multi-Evidence Validation

17,869 human proteins, 14 dietary flavonoids, and 1,496 FDA-approved drugs are connected in a master network of 278,768 interactions. The dataset, created by Koyo Fujisaki and last updated in April 2026, quantitatively predicts flavonoid therapeutic properties, supported by computational, experimental, and epidemiological evidence. Predictions are translated to 506 foods, yielding 685 food-therapeutic combinations.

Use Cases

• Predicting multi-target therapeutic effects of dietary flavonoids based on protein interaction networks.
• Validating computational predictions of flavonoid bioactivity using experimental cytotoxicity data (LC50 values).
• Identifying literature-supported evidence for food-therapeutic associations through systematic PubMed analysis.
• Comparing target protein profiles of flavonoids and FDA-approved drugs for polypharmacology studies.

Strengths

• The master network is quantitatively defined with 278,768 interactions among 17,869 proteins, 14 flavonoids, and 1,496 drugs.
• Computational predictions show high experimental validation, with association strength explaining 84% of variance in potency (Pearson r = 0.918).
• Framework integrates multiple evidence tiers: computational, experimental (cancer cell models), and epidemiological (PubMed analysis).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is small (40.8 KB), indicating a limited scope focused on a specific computational framework.

Provenance

Source

Koyo Fujisaki via figshare

Collection Method

Constructed via a multi-tiered network pharmacology framework integrating computational prediction with experimental and epidemiological evidence.

Freshness

Last updated 2026-04-14 04:14:41