From Plant
to Drug Target
— In One Platform
The first cross-tradition platform connecting ethnopharmacological knowledge from Ayurveda, TCM, Unani, and Western herbalism to protein targets mapped via STITCH, ChEMBL, and Open Targets.
Live Platform
Compound-to-target network for Curcuma longa — 10 human protein targets, 3 bioactive compounds
What We've Built
Infrastructure for botanical drug discovery
Cross-Tradition Knowledge
Ethnobotanical evidence from Ayurveda, TCM, Unani, and Western traditions — unified in one queryable graph.
Compound-to-Target Mapping
Bioactive compounds mapped to protein targets via STITCH and ChEMBL — 940K interactions filtered to 155K high-confidence pairs (composite score ≥ 0.7) with Open Targets disease associations.
Drug Discovery Infrastructure
Druggability scores, scaffold analysis, and TDC ADMET predictions (42 endpoints, Therapeutics Data Commons) + Chemprop GNN models — structured for downstream hit prioritization.
145 GB · 15 Years · 24.4M Literature Links
The Data Moat
Cross-tradition pharmacological datasets don't exist off the shelf. We've spent 15 years assembling 768K structurally catalogued compounds across four medical traditions. Of those, 318K are profiled with Lipinski drug-likeness (MW ≤ 500, LogP ≤ 5, HBD ≤ 5, HBA ≤ 10), ADMET predictions, and protein target annotations — the screened subset the platform operates on.
24.4 million literature linkages connecting plants, compounds, and protein targets across four pharmacological traditions. Disease associations draw on Open Targets scored links (0–1 scale); we surface only associations above 0.2, filtering out purely inferred CTD links. Compound-target-disease graphs that no public database covers end-to-end — enabling queries like: which protein targets are shared between Ashwagandha (Ayurveda) and Reishi (TCM)?
Validation in Progress
Computationally predicted compound–target pairs from the 155K high-confidence dataset are in early wet-lab validation. Validated discovery case studies will be published here as results emerge.
Integrated Sources
Chemistry & Interactions
Traditional Medicine
Pathways & Taxonomy
How It Works
From ethnobotany to target identification
Traditional Knowledge
Ethnobotanical records across 4 medical traditions, curated and cross-linked by plant species.
Compound Analysis
Bioactive compound profiling with TDC ADMET (42 endpoints), Lipinski drug-likeness, and Chemprop GNN target predictions per plant.
Target Discovery
Protein target networks with disease associations, expression data, and clinical pathway linkages.
Market Opportunity
The gap is infrastructure. That's what we built.
Why hasn't this been built sooner? Integrating four distinct medical traditions with modern cheminformatics requires 15 years of manual, domain-specific curation — that's the moat.
See the investment thesis