Our Methodology

Our Mission

PepCodex exists to make peptide research accessible and understandable. We synthesize scientific literature so you can make informed decisions based on evidence, not marketing claims.

We do not provide medical advice, dosing protocols, or sourcing information. Our role is to organize and present what the research says—and critically, what it doesn't say.

Source Selection

We prioritize primary sources in this order:

1. Peer-reviewed human studies — Randomized controlled trials (RCTs), observational studies, case series
2. Clinical trial registries — ClinicalTrials.gov, EU Clinical Trials Register, WHO ICTRP
3. Regulatory documents — FDA, EMA, MHRA approval documents, safety warnings
4. Preclinical studies — Animal and in vitro research (always clearly labeled)
5. Systematic reviews and meta-analyses — When available from reputable sources

Global Coverage

We search beyond English-language databases to include international research. This includes:

• PubMed / MEDLINE (global, primarily English)
• Europe PMC (European focus, open access flagging)
• Russian databases (eLIBRARY.ru, CyberLeninka) when relevant
• Regional registries and regulatory bodies

For non-English sources, we provide translated titles and note the original language.

Evidence Grading

Every peptide dossier includes an evidence strength badge. Here's what each level means:

Citation Standards

Every factual claim in our dossiers is backed by at least one citation. We follow these rules:

• Citations link to the original source (PubMed, DOI, trial registry)
• We note whether sources are open access or paywalled
• We never cite secondary sources when the primary is available
• If we make an inference, we label it clearly as an inference with our confidence level

What We Don't Do

To maintain trust and avoid harm, we explicitly exclude certain content:

• No dosing or protocols — We don't provide instructions on how to use any peptide
• No sourcing information — We don't recommend where to buy anything
• No medical advice — We're researchers, not doctors
• No overclaims — We don't say something is "proven" unless the evidence is overwhelming

How We Update

Peptide research evolves. Our dossiers are living documents:

• Each page shows a "Last Updated" date
• Major updates (new trials, safety signals) trigger revisions
• We maintain a changelog for significant changes

How We Rate Peptides

Every compound receives two independent scores: an Evidence Score (how well-proven it is) and an Effectiveness Score (how large the demonstrated effect is). They are shown together — a compound can be highly effective but lightly proven, or modestly effective but definitively proven. Every criterion is grounded in established evidence-grading frameworks (GRADE, Oxford CEBM, Cochrane Risk of Bias, Bradford Hill, FDA Real-World Evidence), and the scores reflect the science — not a compound's legal or regulatory status.

Evidence Score — the five dimensions

Effectiveness Score

Scored separately, this measures the magnitude of the demonstrated effect for a compound's primary use — graded from effect-size statistics (Cohen's d, NNT) and clinical meaningfulness (MCID), relative to placebo and the best existing option. It carries one of three bases: clinically demonstrated (from human trials), community-reported (from consistent real-world use, capped and clearly labelled when no clinical data exists), or Not Established. It describes effect size — not a recommendation, endorsement, or safety claim.

Evidence composite

The Evidence Score is a weighted average out of 100: (Research Depth × 0.30) + (Mechanism × 0.20) + (Plausibility × 0.20) + (Global Coverage × 0.15) + (Community Experience × 0.15). Research depth is the single heaviest factor, but mechanism, plausibility and real-world use together let a sound, widely-used compound score fairly even without large Western trials.

Score Thresholds

Feedback & Corrections

If you find an error, outdated information, or have a source we should include, please contact us. We take corrections seriously and credit contributors.