Hugging Face's logo

Datasets:
wirthal1990-tech
/
USDA-Phytochemical-Database-JSON

Tasks:
Tabular Classification
Text Classification
Modalities:
Tabular
Text
Formats:
json
Languages:
English
Size:
< 1K
Tags:
phytochemical
ethnobotany
drug-discovery
pubmed
chembl
clinical-trials
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
USDA-Phytochemical-Database-JSON / README.md
Alexander Wirth
fix: update all files to v2.2 — 10 fields, 5 sources, correct pricing, new DOI, 10-field sample
802ff48
preview code
|
raw
history blame contribute delete
14.7 kB
metadata 
license: cc-by-4.0
task_categories:
  - tabular-classification
  - text-classification
language:
  - en
pretty_name: USDA Phytochemical Database — Enriched v2.2
size_categories:
  - 10K<n<100K
tags:
  - phytochemical
  - ethnobotany
  - drug-discovery
  - pubmed
  - chembl
  - clinical-trials
  - patents
  - pubchem
  - smiles
configs:
  - config_name: sample
    data_files:
      - split: sample
        path: ethno_sample_400.json

Full dataset available: Single €699 · Team €1,349 · Enterprise €1,699. → ethno-api.com

USDA Phytochemical & Ethnobotanical Database — Enriched v2.2

The only phytochemical dataset combining USDA botanical records, PubMed citation counts, ClinicalTrials.gov study counts, ChEMBL bioactivity scores, USPTO patent density, and PubChem CID/SMILES — in production-ready JSON + Parquet.

License: CC BY 4.0 Sample Full Dataset Format HuggingFace DOI

Free 400-Row Sample ↓ · Single €699 → · Team €1,349 → · Enterprise €1,699 →

Enrichment status (March 2026): All five enrichment layers (PubMed, ClinicalTrials.gov, ChEMBL, PatentsView, PubChem) are complete and final. The free 400-row sample contains real enrichment values.

Records Compounds Species Enrichment Layers
76,907 24,746 2,313 5

Data Quality: Dataset was audit-validated on 2026-03-16. Original 104,388 records cleaned to 76,907 by removing macronutrients (WATER, GLUCOSE etc.) and exact duplicates. [Audit report available on request.]

Schema (v2.2)

Column Type Nulls Description
chemical string 0% Standardised compound name (USDA Duke's nomenclature)
plant_species string 0% Binomial Latin species name
application string ~50% Traditional medicinal application (e.g. "Antiinflammatory")
dosage string ~87% Reported dosage, concentration, or IC50 value
pubmed_mentions_2026 int32 0% Total PubMed publications mentioning this compound (March 2026 snapshot)
clinical_trials_count_2026 int32 0% ClinicalTrials.gov study count per compound (March 2026)
chembl_bioactivity_count int32 0% ChEMBL documented bioactivity measurement count
patent_count_since_2020 int32 0% US patents since 2020-01-01 mentioning compound (USPTO PatentsView)
pubchem_cid int64 ~58% PubChem Compound ID (CID) — resolved via PubChem PUG REST (March 2026)
canonical_smiles string ~58% Canonical SMILES notation — molecular structure from PubChem (42.4% of unique compounds resolved)

Pricing & Licensing

Tier Price Includes Purchase
Single Entity €699 netto JSON + Parquet + SHA-256 Manifest. 1 juristische Person, interne Nutzung. Perpetual license. Buy Now →
Team €1,349 netto Alles aus Single + duckdb_queries.sql (20 Queries, 5 Kategorien) + compound_priority_score.py + 4 Pre-computed Views (Top-500 nach PubMed, Trials, Patent-Dichte, Anti-Inflammatory Panel). Unbegrenzte interne Nutzer einer juristischen Person. Buy Now →
Enterprise €1,699 netto Alles aus Team + snowflake_load.sql + chromadb_ingest.py + pinecone_ingest.py + embedding_guide.md (ClinicalBERT, RAG-Pipelines) + Compound Opportunity Matrix + Clinical Pipeline Gaps CSV + Pre-chunked RAG JSONL. Multi-Entity / Konzernnutzung, interne Produktintegration erlaubt. Contact →

Gemäß § 19 UStG wird keine Umsatzsteuer berechnet. Alle Preise netto. One-time purchase — keine Subscription, keine wiederkehrenden Kosten.

Why Not Build This Yourself?

Normalising and cross-referencing 24,746 phytochemicals across multiple authoritative sources is not a weekend project.

Scope Effort Cost @ $85/hr
USDA cleaning + normalization + enrichment + exports + QA 48–60h ~$4,080–$5,100

This dataset: €699. No subscription. No API calls. Download link sent instantly after payment. Valid for 72 hours. See ethno-api.com.

Why This Dataset Exists

Large language models hallucinate botanical taxonomy. A biotech team's RAG pipeline confidently outputting "Quercetin found in 450 species at 2.3 mg/g" sounds plausible — but the real number of species in our data is 215, and dosage varies by three orders of magnitude depending on the plant part.

The raw USDA Dr. Duke's database is spread across 16 relational tables. Joining them correctly requires understanding non-obvious foreign keys, handling >40% null values in application fields, and normalising species names against accepted binomial nomenclature. Most teams give up after a week.

Quickstart

Python — Load 400-row sample 

import pandas as pd

url = "https://raw.githubusercontent.com/wirthal1990-tech/USDA-Phytochemical-Database-JSON/main/ethno_sample_400.json"
df = pd.read_json(url)
print(f"{df.shape[0]} records, {df['chemical'].nunique()} unique compounds")
df.head()

PyArrow — Parquet (full dataset, after purchase)

Download link delivered instantly after payment (valid 72h). See ethno-api.com.

import pyarrow.parquet as pq

table = pq.read_table("ethno_dataset_2026_v2.2.parquet")
print(f"Schema: {table.schema}")
print(f"Rows: {table.num_rows}  Memory: {table.nbytes / 1e6:.1f} MB")

DuckDB (analytical queries — sample included) 

import duckdb

result = duckdb.sql("""
    SELECT
        chemical,
        MAX(pubmed_mentions_2026)       AS pubmed_score,
        MAX(clinical_trials_count_2026) AS trial_count,
        MAX(chembl_bioactivity_count)   AS bioassays,
        COUNT(DISTINCT plant_species)   AS species_count
    FROM read_json_auto('ethno_sample_400.json')
    GROUP BY chemical
    ORDER BY trial_count DESC
    LIMIT 20
""")
result.show()

HuggingFace Datasets 

from datasets import load_dataset

ds = load_dataset(
    "wirthal1990-tech/USDA-Phytochemical-Database-JSON",
    split="sample",
    trust_remote_code=False
)
df = ds.to_pandas()
print(f"Records: {len(df)} | Columns: {list(df.columns)}")
df.head()

Note: The split="sample" loads ethno_sample_400.json (400 rows, 10 columns). The full 76,907-row dataset is available at ethno-api.com.

Sample Record

Below is a real record from the dataset — QUERCETIN, one of the most-studied plant compounds:

{
  "chemical": "QUERCETIN",
  "plant_species": "Drimys winteri",
  "application": "5-Lipoxygenase-Inhibitor",
  "dosage": "IC50 (uM)=4",
  "pubmed_mentions_2026": 31310,
  "clinical_trials_count_2026": 81,
  "chembl_bioactivity_count": 2871,
  "patent_count_since_2020": 73,
  "pubchem_cid": 5280343,
  "canonical_smiles": "C1=CC(=C(C=C1C2=C(C(=O)C3=C(C=C(C=C3O2)O)O)O)O)O"
}

All 76,907 records contain all 10 schema fields. The 4 integer enrichment columns (pubmed_mentions_2026, clinical_trials_count_2026, chembl_bioactivity_count, patent_count_since_2020) are always non-null; pubchem_cid and canonical_smiles are filled for 42.4% of unique compounds (10,484 of 24,746 resolved via PubChem PUG REST); application (50% null) and dosage (87% null) reflect USDA source gaps. Unresolved compounds are phytochemical trivial names, mixture descriptions, or non-specific ethnobotanical terms not indexed in PubChem by name. The free 400-row sample contains real, final enrichment values across all five layers.

File Manifest

File Size Format Access
ethno_sample_400.json 131 KB JSON Free (this repo)
ethno_sample_400.parquet 20 KB Parquet Free (this repo)
quickstart.ipynb 9 KB Notebook Free (this repo)
ethno_dataset_2026_v2.2.json ~25 MB JSON Included in all tiers
ethno_dataset_2026_v2.2.parquet ~1.2 MB Parquet Included in all tiers
MANIFEST_v2.json (SHA-256) ~1 KB JSON Included in all tiers
duckdb_queries.sql (20 Queries) ~13 KB SQL Team + Enterprise
compound_priority_score.py ~5 KB Python Team + Enterprise
snowflake_load.sql ~6 KB SQL Enterprise
chromadb_ingest.py ~6 KB Python Enterprise
pinecone_ingest.py ~6 KB Python Enterprise
embedding_guide.md ~7 KB Markdown Enterprise

Data Sources & Provenance

All enrichment layers are derived from authoritative, publicly accessible scientific databases and represent a March 2026 snapshot.

Source Snapshot What it contributes
USDA Dr. Duke's Phytochemical and Ethnobotanical Databases 2026 Canonical plant–compound–application baseline across 2,313 species
NCBI PubMed March 2026 Compound-level publication evidence score
ClinicalTrials.gov March 2026 Compound-level clinical research activity score
ChEMBL March 2026 Compound-level bioactivity measurement depth
USPTO PatentsView March 2026 Compound-level commercial IP activity score
PubChem March 2026 PubChem CID + Canonical SMILES molecular structure notation

Enrichment methodology is documented in METHODOLOGY.md. Source code is available to Enterprise license holders upon request under NDA.

Use Cases

  • RAG Pipelines — Ground LLM responses with verified phytochemical data. Each record has a PubMed evidence score — use it to weight retrieval results and filter hallucinations.
  • Drug Discovery — Prioritise natural product leads by combining PubMed citations, clinical trial presence, ChEMBL bioactivity depth, and patent landscape. One query replaces weeks of manual lit review.
  • Market Intelligence — Patent density score reveals which compounds are attracting commercial investment. Cross-reference with clinical trials to identify underexplored compounds with IP whitespace.
  • Academic Research — Pre-computed evidence scores save months of PubMed searching. The BibTeX citation block below makes this dataset citable in peer-reviewed publications.

Dataset Versions

Version Records Schema Status
v1.0 104,388 5 columns (USDA baseline) Deprecated
v2.0 76,907 8 columns (+ PubMed, ClinicalTrials, ChEMBL, Patents) Deprecated
v2.1 76,907 10 columns (+ PubChem CID, Canonical SMILES) Superseded
v2.2 76,907 10 columns (stereo-prefix CT normalization, corrected SMILES coverage) Current

The free sample (ethno_sample_400.json) uses the v2.2 schema with final enrichment values across all five layers.

Data Attribution

This dataset includes bioactivity count data (chembl_bioactivity_count field) derived from ChEMBL v35, licensed under CC BY-SA 3.0. Buyers who redistribute this field downstream must comply with ChEMBL attribution requirements.

License & Commercial Access

  • Free 400-row sample: CC BY 4.0 — use for evaluation, academic research, and prototyping.
  • Single Entity License — €699 one-time: Buy → — 1 legal entity, internal use, perpetual. No redistribution.
  • Team License — €1,349 one-time: Buy → — all employees of 1 legal entity, unlimited internal users, includes analytics toolkit.
  • Enterprise License — €1,699 one-time: Contact → — multi-entity / group use, internal product integration rights, full RAG integration toolkit.

Gemäß § 19 UStG wird keine Umsatzsteuer berechnet.

Citation

If you use this dataset in your research, please cite:

Wirth, A. (2026). USDA Phytochemical Database — Enriched v2.2 (Sample). Zenodo. https://doi.org/10.5281/zenodo.15083493

@misc{ethno_api_v2.2_2026,
  title     = {USDA Phytochemical \& Ethnobotanical Database --- Enriched v2.2},
  author    = {Wirth, Alexander},
  year      = {2026},
  publisher = {Ethno-API},
  url       = {https://ethno-api.com},
  doi       = {10.5281/zenodo.15083493},
  note      = {76,907 records, 24,746 unique chemicals, 2,313 plant species, 10-column schema with PubMed, ClinicalTrials, ChEMBL, PatentsView, PubChem CID/SMILES enrichment}
}

DOI

Contact

Built by Alexander Wirth · PostgreSQL 15 · Python 3.12 · Hetzner CCX33