From Plant to Drug Target
Automated.

BioPipeline indexes 2,377+ medicinal plant species with 104,000+ phytochemical compounds from Dr. Duke's Phytochemical and Ethnobotanical Databases. This includes well-studied species like Curcuma longa (turmeric), Cannabis sativa, Panax ginseng, Zingiber officinale (ginger), and thousands more traditional medicine plants.

Most runs complete in 2-5 minutes depending on the number of targets selected. Simple runs (1-3 targets) finish in under 3 minutes, while comprehensive runs with 10+ targets may take 5-7 minutes. Pro and Max users get priority queue access for faster processing during peak hours.

Max and Enterprise plans support custom uploads. You can upload protein structures in PDB format and compound libraries in SDF, MOL2, or SMILES format. The SMILES Dock tool (available on all plans) allows single-compound docking against any PDB structure via PDB ID or file upload.

Results can be exported in multiple formats:

• CSV — Spreadsheet data with all metrics (all plans)
• JSON — Structured data for programmatic access (all plans)
• PDF — Professional report with visualizations (Pro+)
• Excel — Formatted workbook with multiple sheets (Pro+)
• SDF — 3D structures for further analysis (Max+)
• PDBQT — Docked poses for validation (Max+)

Our ADMET predictions use established RDKit descriptors and Lipinski's Rule of Five, which are industry-standard computational methods. However, these are computational predictions, not experimental data. For regulatory submissions or clinical decisions, we strongly recommend experimental validation using in vitro assays (Caco-2, MDCK, microsomal stability, hERG binding, etc.).

Yes! Academic institutions receive 50% off Pro and Max plans. Simply contact us at billing@biopipeline.online with your .edu email address and proof of affiliation (student ID, faculty page, or institutional email). The discount applies to annual subscriptions and includes all features of the respective tier.

Max plan includes full REST API access for programmatic submissions and result retrieval. Enterprise plans support custom integrations, webhooks, batch processing, and dedicated endpoints. API documentation is available in-app after upgrading. Rate limits: Max (1000 req/day), Enterprise (custom).

We use Groq's Llama 3 model (70B parameters) to analyze docking results, ADMET profiles, and compound properties. The AI generates human-readable summaries, identifies promising lead candidates, suggests experimental validation strategies, and highlights potential drug-drug interactions or safety concerns. All AI insights include confidence indicators and should be reviewed by qualified researchers.

AutoDock Vina typically achieves ±2 kcal/mol accuracy compared to experimental values for well-prepared systems. Predictions are most reliable for:

• High-resolution protein structures (< 2.5 Å resolution)
• Compounds within drug-like chemical space
• Known binding sites with experimental validation

Predictions are less reliable for flexible proteins, allosteric sites, or compounds with unusual chemistry. Always validate top hits experimentally before investing in synthesis or further development.

Your pipeline runs and results are private by default and visible only to your account. We do not share, sell, or train AI models on your proprietary data. You can permanently delete runs from your history at any time. Enterprise customers can request data residency options and dedicated infrastructure. See our Privacy Policy for complete details.