We map 6,500+ medicinal plants across multilingual pharmacopeias β TCM, Ayurveda, Kampo, Unani, and Western herbal β to validated molecular endpoints using PubMed synthesis and computational pharmacology. Cross-database synergy optimisation unlocks Eastern sources for Western R&D pipelines. Deliverables formatted for direct R&D and regulatory handoff.
Each engagement produces a structured research artifact β not a literature dump β designed for direct use by formulators, R&D directors, and regulatory teams.
Quantitative scoring of 6,500+ medicinal plants against your specified molecular endpoints. Coverage = validated hits / total targets Γ evidence weight. Stratified across ETCM 2.0, AyurvedaDB, IMPPAT 2.0, and DrugBank with InChIKey normalisation across all sources.
Multi-herb combination analysis using target complementarity matrices across all 28 databases. Loewe additivity surface modelling for orthogonal mechanisms. Constraint-aware ranking: dosage form, solubility, regional regulations, exclusion lists. Output: 3β5 formula candidates with pairwise interaction scores.
Systematic PubMed extraction across 1.7M+ pharmacology records. Structured search: [compound] AND [target] AND (IC50 OR Kd OR "network pharmacology" OR "molecular docking"). Evidence hierarchy: L1 direct binding β L4 traditional use. Full citation matrix for regulatory documentation support.
GRAS/Novel Food status verified per herb. EMA, FDA 21 CFR, PMDA, TGA, and ASEAN compliance matrices applied. Extraction ratio data, stability flags, elemental impurity screening per ICH Q3D referenced. All outputs structured for documentation use β not a substitute for clinical or laboratory confirmation.
Graph neural networks trained on ChEMBL + PubMed for target prediction. XGBoost ranking for coverage optimisation. SHAP explainability for every recommendation. Monte Carlo simulations generating synergy confidence intervals. DisGeNET integration: 1.13M gene-disease associations filtered to client targets.
87% of published pharmacology studies use male cell lines or animal models. We apply systematic re-weighting across all body systems via EDN 3.2, CycleBase 3.0, and female-stratified CYP450 polymorphism data β correcting for metabolism, serotonin signalling, cortisol regulation, immune cycling, and neurotransmitter density variations. Every compound re-scored. Not reproductive health only β full female physiology.
The NIH only mandated inclusion of female cell lines in federally funded research in 2016. Prior to this β and in the majority of legacy data still in use β target validation, dose-response characterisation, and compound profiling was conducted exclusively in male models.
The consequence: female-specific differences across CYP450 metabolism, serotonin signalling, cortisol regulation, immune response cycling, and neurotransmitter receptor density remain systematically under-characterised in every major public pharmacology database. This affects all formulation decisions β not just hormone-targeted products.
Luminae applies targeted corrections through integration of endocrine disruption networks (EDN 3.2), menstrual cycle transcriptomics (CycleBase 3.0), and sex-stratified clinical endpoints β producing target profiles reflecting actual female physiology across all body systems.
All outputs are computational research documentation for R&D use. Laboratory and clinical confirmation is required before any product development decision.
| Body System | Male Study Bias | Luminae Coverage |
|---|---|---|
| CYP450 metabolism | 92% | 94% |
| Serotonin (HTR1A/2A) | 79% | 88% |
| Cortisol axis (HPA) | 83% | 90% |
| Immune cycling (IL-6/TNF-Ξ±) | 85% | 91% |
| Neurotransmitter density | 76% | 87% |
| ESR1 / ESR2 signalling | 68% | 89% |
| GNRHR / FSHR axis | 54% | 82% |
Compound records standardised using InChIKey normalisation across all sources. Target resolution to UniProt canonical sequences. Daily ETL pipeline. Non-English sources (Mandarin, Sanskrit, Japanese) processed via multilingual NLP β unlocking Eastern databases inaccessible to English-only tools.
1,247 herbs Β· 8,471 compounds Β· 497 targets. Primary herb-compound-target mapping. Mandarin NLP extraction.
1,892 substances from Li Shizhen's encyclopedia. Fu Qingzhu: 100+ formulas for female physiological balance. Mandarin source.
892 herbs Β· 4,010 phytochemicals Β· 1,226 validated targets. Largest computational Ayurveda resource.
Female physiology chapters. Shatavari, Ashwagandha ERΞ±/PR phytoestrogen data. Sanskrit source processed.
212 Kampo formulas Β· 2,847 constituents Β· RCT registry linked to compound data. Japanese source.
1,496 botanical species Β· 14,827 compounds. EMA community herbal monograph compliance status included.
320 Unani materia medica Β· 3,200+ phytochemicals. WHO traditional medicine monograph cross-referenced.
2.1M bioactivity assays. IC50/Kd/Ki gold-standard binding data for herbal target cross-validation.
1.13M gene-disease associations filtered to client targets. 35,128 natural compounds Γ 6,083 targets ML-predicted.
26,000+ TCM herb-target mappings. Clinical study integration linking traditional use to validated molecular pathways.
Endocrine disruption network Β· 2,847 endocrine targets. Menstrual cycle phase transcriptomics Β· female CYP450 polymorphisms.
1.7M+ pharmacology records. Daily ETL. Structured extraction: binding affinities, pathways, clinical endpoints.
Every engagement follows a reproducible, auditable pipeline designed to meet pharmaceutical-grade documentation standards. Each phase output is version-controlled and referenced in the final deliverable.
Client brief parsed to UniProt/Entrez canonical IDs. Female physiology weighting applied across all systems: CYP450 Γ1.9, serotonin Γ1.6, immune cycling Γ1.4, cortisol Γ1.2.
4-way intersection: ETCM 2.0 Β· AyurvedaDB Β· SuperPred Β· DrugBank. Consensus interactions only. Coverage score = Ξ£ weighted hits / total targets Γ evidence level multiplier.
Pairwise herb complementarity scored. Loewe additivity modelling for mechanism overlap. Multi-objective optimisation: maximise coverage, minimise antagonism (<0.3 threshold), apply client constraints.
GRAS/Novel Food status verified. EU, US, ASEAN, JP restriction matrices applied. Extraction ratios, stability flags, and elemental impurity data referenced per ICH Q3D.
All associations validated against PubMed full-text. Evidence levels assigned L1βL4. Traceability matrix: PubMed ID β compound β target β herb β formula candidate.
Indicative output β PCOS / hormonal balance
| Target | Weight | Coverage |
|---|---|---|
| ESR1 / ESR2 | Γ1.8 | 89% |
| AR (androgen receptor) | Γ1.5 | 82% |
| INSR / IGF1R | Γ1.4 | 78% |
| PTGS2 / IL-6 | Γ1.4 | 91% |
| SHBG | Γ1.2 | 71% |
All outputs in editable Excel + PDF. Structured for direct handoff to R&D teams, contract manufacturers, and regulatory affairs consultants. Research documentation only β laboratory and clinical validation required before product decisions.
Weighted molecular endpoint table with evidence strength scores and female physiology correction factors. Excel + PDF.
Full 6,500+ plant ranking against your target set. Coverage heatmap. Hit distribution by database source.
3β5 constraint-compliant herb combinations. Coverage scores, synergy ratings, trade-off notes.
PubMed ID β compound β target β herb. Evidence level per association. Exportable CSV for regulatory documentation use.
GRAS, Novel Food, EMA monograph status per herb. Regional restriction flags. Elemental impurity references.
Feature importance for all ranked recommendations. Audit trail linking GNN predictions to primary literature evidence.
Send your indication, primary molecular targets, and formulation constraints. Scope confirmed within 24 hours. Full analysis delivered within 48β72 hours.
Send Analysis Request