Methodology | Peptide Database

Data Sources

Each compound profile is built from multiple evidence tiers. We distinguish between human clinical data, animal studies, in-vitro research, and mechanism-based theoretical analysis. Every claim is tagged with its evidence level so you know what you're looking at.

Clinical

Human trials, FDA-approved drug labels, published clinical data. Compounds like semaglutide, testosterone, and metformin have extensive clinical evidence supporting their profiles.

Preclinical

Animal studies and well-powered preclinical trials. Peptides like BPC-157, TB-500, and most growth hormone secretagogues fall into this category with strong but non-human evidence.

In Vitro & Emerging

Cell studies and early-stage research. Newer peptides and bioregulators often have limited but promising data. We flag these clearly so you can calibrate your expectations.

The Interaction Engine

Our interaction checker doesn't just look up a table. It runs a 3-tier pharmacological inference engine that evaluates every compound pair through multiple resolution layers, producing a confidence-scored assessment even when no direct study of that specific pair exists.

Explicit Research Data

Over 850 interaction entries documented from published research, clinical guidelines, and established protocols. Each entry includes the interaction status (synergistic, compatible, monitor, avoid, or timing-required) and a specific note explaining why.

Mechanism-Based Inference

When no direct study exists for a pair, the engine evaluates pharmacological tags assigned to each compound: receptor targets, signaling pathways, organ load profiles, and safety flags. A rule engine with 48 pharmacological rules checks for known interaction patterns like shared receptor competition, additive toxicity, complementary healing pathways, or hormonal conflicts.

Confidence is adjusted based on the evidence quality of each compound's pharmacological profile. Two clinically-studied compounds produce a higher-confidence inference than two compounds with only theoretical data.

Theoretical Fallback

For pairs where no rules match, the engine compares organ system overlap. If both compounds affect the same organs, it recommends standard monitoring. If they act on entirely different systems, it notes no known conflicts. This ensures every pair gets at least a baseline assessment rather than "no data."

Pharmacological Profiling

Every compound is tagged with a structured pharmacological profile extracted from its mechanism of action, clinical data, side effect profile, and contraindication data. These tags power the interaction engine and stack analysis.

Targets

Receptor and enzyme targets — androgen receptor, GLP-1 receptor, aromatase, PDE5, AMPK, ghrelin receptor, and dozens more. Used to detect receptor competition and complementary mechanisms.

Pathways

Signaling cascades and biological pathways — protein synthesis, GH-IGF1 axis, angiogenesis, autophagy, HPG axis, wound healing. Used to identify synergistic and antagonistic combinations.

Safety Flags

Risk markers like hepatotoxic, HPTA-suppressive, cardiotoxic, serotonergic, estrogenic. The engine flags when multiple compounds in a stack share safety risks — cumulative toxicity that pairwise checks alone would miss.

Stack-Level Analysis

Beyond pairwise interactions, the engine performs cumulative stack analysis. When you check 3+ compounds together, it doesn't just check every pair — it also evaluates:

Organ load accumulation — each compound declares which organs it stresses (liver, heart, kidneys, etc.) and at what level. The engine sums these across your entire stack to flag when cumulative organ stress exceeds safe thresholds.

Safety flag aggregation — if 3 compounds in your stack are all hepatotoxic, that's a stack-level danger that no individual pair check would catch.

Overall risk scoring — combines pairwise interactions, organ loads, and safety flags into a single risk level (low, moderate, elevated, high) with specific warnings.

Quality Standards

Transparency

Every interaction shows its confidence percentage, resolution tier, and source. You always know whether an assessment comes from documented research or pharmacological inference.

Conservative Defaults

When uncertain, the engine errs on the side of caution. A compound flagged in contraindications as affecting the liver gets the hepatotoxic safety flag even if the risk is low. Better a false "monitor" than a false "all clear."

Continuous Updates

Compound profiles, pharmacological tags, and interaction rules are updated as new research emerges. Each profile tracks its research status from "limited research" through "FDA approved."

By the Numbers

Peptide Database is an educational and research resource. It does not provide medical advice, diagnose conditions, or recommend specific protocols. All interaction assessments should be verified with a qualified healthcare professional before making decisions about compound use.