Research Peptide and Compound Library

Research compounds are often described with a mix of laboratory findings, animal data, human studies, regulatory records, marketing language, and speculation. Those are not interchangeable forms of evidence.

The Serum Research Library organizes information by material identity, study model, result, limitation, and regulatory status. Its purpose is to help readers answer four separate questions:

1. What exact substance or preparation does a source describe?
2. What question did the study or official record address?
3. What result was actually measured?
4. What does that result not establish?

This library is educational. It does not provide medical advice, treatment recommendations, dosing instructions, reconstitution guidance, or directions for human or veterinary use. A library record does not authenticate, validate, or establish the suitability of a catalog material.

Sources checked: August 31, 2026. Questions or corrections can be sent through the Serum Research contact page.

Editorial responsibility: Serum Research. This page was prepared from the primary and official sources listed below. No independent scientific, regulatory, or clinical review is claimed.

Explore the library: Use the Research Evidence Guides for study design and result interpretation, the Research Material Identity Guides for substances, forms, and analytical boundaries, and the Research Study Records for individual trial records.

Browse peptide research questions and evidence

Evidence-reading principles used in this library

The sections below are concise, self-contained reading principles used across the library. The dedicated public guides below connect each reading principle to its full source-first checklist.

How Serum Research reviews evidence

The library review method separates source selection, evidence tiers, identity checks, limitations, correction handling, and the difference between a study result and a product claim.

How to read a peptide or research-compound study

A practical research-study reading framework starts with the exact material, model or population, comparator, prespecified primary outcome or outcomes, success rule, effect estimate, uncertainty, missing data, funding, and reporting limits before accepting a headline conclusion.

In vitro vs animal vs human evidence

The in vitro, animal, and human evidence guide separates the questions answered by cell, animal, observational-human, and interventional-human studies. It marks where each evidence model stops so a mechanistic or preclinical result is not silently converted into a human outcome.

What a negative or null result means

The negative and null result guide explains how a non-significant result can reflect a small effect, an imprecise estimate, low information, a design limitation, or evidence compatible with no meaningful difference. Confidence intervals, prespecified outcomes, multiplicity, and missing data all matter when interpreting it.

How to read confidence intervals, p values, and multiplicity

The confidence intervals, p values, and multiplicity guide shows how effect size, precision, a threshold comparison, and the number of tested hypotheses answer different questions. It explains how to read them together and why nominal, exploratory, and multiplicity-adjusted results must remain distinct.

Trial estimands, intercurrent events, and missing data

An estimand defines the treatment effect a trial is trying to estimate. The trial estimands, intercurrent events, and missing-data guide separates that target from the statistical estimator, events such as treatment discontinuation or rescue intervention, missing observations, analysis populations, and sensitivity analyses.

Surrogate endpoints versus clinical outcomes

A biomarker, image, physiological measure, or other surrogate does not automatically show how people feel, function, or survive. The surrogate endpoints and clinical outcomes guide explains why validation is context-specific, and why a surrogate result or accelerated-approval record does not transfer to another material.

Randomization, masking, and control groups

Randomization, allocation concealment, masking, and comparators address different risks of bias. The randomization, masking, and control-group guide distinguishes concurrent controls from external or historical controls and separates no-treatment groups from protocol-defined usual care.

Absolute versus relative effects and baseline risk

A relative percentage can sound large while the absolute difference is small, or the reverse. The absolute-versus-relative-effects guide keeps the outcome, group risks, denominator, time frame, baseline risk, effect scale, and uncertainty visible together.

Subgroup analyses and interaction tests

A favorable result in one subgroup and a non-significant result in another do not, by themselves, prove that the treatment effect differs. The subgroup-analysis and interaction-test guide explains how prespecification, multiplicity, plausibility, precision, and replication affect that interpretation.

Adverse events, serious adverse events, and tolerability

Seriousness, severity, causality, participant counts, event counts, discontinuation, exposure time, and tolerability are separate fields. The adverse-event and tolerability guide explains why a brief safety summary cannot replace the study’s collection methods, denominators, and detailed tables.

Systematic reviews, meta-analysis, heterogeneity, and publication bias

A pooled estimate inherits the eligibility decisions, biases, outcome definitions, and missing evidence of its included studies. The systematic review and meta-analysis guide explains how to assess protocols, searches, effect models, heterogeneity, sensitivity analyses, small-study effects, publication bias, and certainty without treating a forest plot as a verdict.

Trial registration, protocols, analysis plans, and outcome switching

The ClinicalTrials.gov and FDA status-check guide separates a registry entry, protocol, statistical analysis plan, results record, journal article, and product-specific approval record. Their dates, prespecified outcomes, amendments, analyses, and reported results can be compared using the trial registration and outcome-switching guide, without treating registration, an identifier, or a database listing as proof of completion, quality, success, or approval.

Composite endpoints, co-primary endpoints, and components

A composite combines events, while co-primary endpoints create more than one required primary test. The endpoint family, component definitions, event counts, effect estimates, uncertainty, multiplicity rule, and clinical importance must remain visible so a favorable composite is not mistaken for uniform benefit across its components. See the composite and co-primary endpoint guide for a structured reading checklist.

Noninferiority, equivalence, and superiority trials

These designs answer different questions and use different decision rules. Margins, assay sensitivity, constancy assumptions, confidence-interval boundaries, and analysis populations matter, and failure to show superiority does not establish equivalence or noninferiority. See the noninferiority, equivalence, and superiority guide for the decision-rule checklist.

Dose-response and exposure-response evidence

Dose assigned by protocol and exposure measured in a participant are related but not interchangeable. This section separates dose groups, observed exposure, modeled relationships, endpoint selection, multiplicity, uncertainty, and material/formulation identity so an observed trend is not converted into a universal regimen or product claim. See the dose-response and exposure-response guide for a structured reading checklist.

Clinical trial phases describe programs, not certainty

Phase labels describe common development objectives; they are not evidence grades. Phase must be separated from the study question, design, population, comparator, endpoints, duration, result, and accumulated evidence. Registered, completed, or Phase 3 does not itself mean favorable, approved, safe, or broadly applicable. See the clinical trial phases and evidence guide for the distinctions between development labels, study objectives, post-approval research, and product-specific approval.

Risk of bias, reporting quality, and evidence certainty

Risk of bias for a result, completeness of reporting, precision, directness, and certainty across a body of evidence are different judgments. These layers should remain distinct without turning a reader-level appraisal into a validated scoring tool or assuming that randomization removes every source of bias. See the risk-of-bias and evidence-certainty guide for result-specific appraisal, reporting limitations, and outcome-specific certainty judgments.

Intention-to-treat, full analysis, and per-protocol sets

The intention-to-treat principle, an operational full-analysis set, missing-data method, and per-protocol set are not synonyms. Post-randomization exclusions matter, analysis sets must be tied to the trial question, and noninferiority and equivalence settings require particular care. Read the intention-to-treat and per-protocol guide for estimand alignment, justified exclusions, missing data, and noninferiority caveats.

Crossover trials: sequence, period, and carryover

A randomized crossover trial assigns participants to an intervention sequence, making period effects, washout, carryover, retention, and within-participant analysis central to interpretation. Read the crossover trials guide for a structured appraisal of sequence allocation, period-specific outcomes, paired analysis, missing observations, and the limits of carryover testing.

Statistical significance, meaningful change, and responders

A p value does not measure effect size, clinical importance, benefit probability, or the probability that a hypothesis is true. Group means, within-person change, context-specific meaningful-change thresholds, responder definitions, uncertainty, multiplicity, and missing data must remain separate. Read the meaningful-change and responder guide for context-specific score interpretation, justified thresholds, and between-group effects.

What establishes peptide sequence and structural identity?

A short name or nominal residue string does not settle sequence direction, stereochemistry, terminal chemistry, modification sites, cyclization, or disulfide connectivity. The peptide sequence and structural identity guide explains which structural fields must agree before two peptide records can be treated as the same defined entity, and why a database identifier still does not authenticate a lot.

Peptide free base, salt form, counterions, and labeled mass

A peptide active moiety may appear as a free base, acetate-associated material, or another salt or counterion context. The free-base, salt, and counterion guide explains why chemical form, stoichiometry, hydration or solvation, formula mass, active-moiety mass, and declared container amount must remain separate so one label basis is not silently substituted for another.

Milligrams, concentration, and International Units

Milligrams describe mass, concentration requires a stated denominator, and International Units describe biological activity against a defined reference and assay system. The mass, concentration, and International Units guide explains why those quantities answer different questions and why a product-specific relationship cannot be reused as a universal conversion.

Active ingredient, formulation, excipients, and delivery system

An active-ingredient name does not establish concentration, excipients, release behavior, container, delivery system, stability, exposure, or finished-product approval. The formulation, excipients, and delivery-system guide explains why evidence for a defined formulation must remain separate from another preparation or research material that shares only an ingredient name.

How to check ClinicalTrials.gov and FDA status

Trial registration, completion, posted results, FDA review, approval, an NDC, and a UNII are different records. Status should be checked in primary records without treating an identifier or registry entry as endorsement.

FDA-approved drug vs research material

FDA approval belongs to a named finished product and approved application, with controlling product information and current labeling. A separately sold catalog material does not inherit that approval merely because it shares an ingredient name.

What quality-testing methods can and cannot show

Start with the peptide quality-testing guide and the lot-linked certificate guide to separate identity, purity, quantity, sterility, endotoxin state, stability, and lot linkage.

Identity, chromatographic purity, quantitative content, sterility, endotoxin state, residual solvents, stability, and lot linkage are separate questions. Common analytical records can address only the particular method, sample, analyte, result, and lot linkage they actually document.

Research-topic coverage

The topic groupings below describe published hubs and planned coverage. They do not imply that the listed materials share a mechanism, effect, safety profile, regulatory status, or appropriate use. No topic hub or compound record is implied to be public unless it is linked from this page.

Growth-hormone and endocrine research

The growth-hormone and endocrine research hub separates identity and evidence for CJC-1295 with and without DAC, ipamorelin, tesamorelin, hexarelin, somatropin, IGF-1 LR3, kisspeptin, hCG, and menotropins. It distinguishes releasing-hormone analogues, secretagogues, recombinant hormones, growth-factor analogues, isoforms, and activity-standardized mixtures.

Metabolic, mitochondrial, and redox research

The metabolic, mitochondrial, and redox research hub covers MOTS-c, SS-31/elamipretide, NAD+, L-carnitine, glutathione, retatrutide, tirzepatide, cagrilintide, AOD-9604, 5-Amino-1MQ, and SLU-PP-332. Grouping these records does not turn a mitochondrial peptide, cofactor, amino-acid derivative, receptor agonist, peptide fragment, or small molecule into an interchangeable class.

Melanocortin research

The melanocortin research hub covers afamelanotide/Melanotan I, Melanotan II, bremelanotide/PT-141, and KPV while preserving the sequence, cyclization, receptor, finished-product, and evidence differences among them.

Neurobiology and CNS research

The neurobiology and CNS research hub covers Semax, Selank, DSIP/emideltide, Cerebrolysin, and Epitalon/Epithalon records organized by exact identity, study model, human evidence, null findings, and current regulatory context. The grouping is editorial and does not recommend a compound for sleep, anxiety, cognition, stroke, recovery, or longevity.

How published study records are organized

The study-record reading primer explains how to assess the test article, study question, population, comparator, outcome hierarchy, effect and uncertainty, missing data, adverse events, sponsor role, registration, publication, and the exact limit on what can be inferred. Individual study records examine one trial or primary paper at a time.

Published study records are not condensed treatment summaries. They will contain no product links and will not transfer a trial result to a catalog material with a similar name.

How compound records are organized

A compound record is published only when it has a distinct identity or question, a primary or official source trail, a non-duplicate evidence summary, visible limitations, a review date, and a real correction route.

Each record follows a consistent structure:

1. Identity and names, the preferred name, recognized synonyms, sequence or structure when an authoritative record supports it, and forms that must remain distinct.
2. Evidence by model, biochemical, cell, animal, observational-human, and interventional-human findings appear separately.
3. Negative, null, mixed, or missing findings, results that limit a claim receive the same prominence as favorable observations.
4. Regulatory or development status, current FDA and ClinicalTrials.gov records are described with an as-of date and their exact scope.
5. What the evidence does not establish, the population, material, formulation, endpoint, duration, and product-equivalence boundaries.
6. Primary and official sources, stable records such as PMID, DOI, NCT, FDA application, current label, or substance identifier where available.

Compound records are released only after independent review of their clean public paths. This hub will list only records that are already public and approved; a private draft is not a hub item.

Evidence models used in the library

A higher evidence tier is not a product rating. It does not erase bias, imprecision, poor reporting, a null primary result, limited generalizability, or a mismatch between the studied material and a commercial sample.

Source policy

The library prioritizes:

• FDA labels, application records, review documents, safety records, and official status pages for U.S. regulatory statements;
• ClinicalTrials.gov records for registered design, recruitment, sponsor-reported results, and trial identifiers;
• peer-reviewed primary studies for experimental findings;
• authoritative identity and reference-standard records where they resolve substance naming; and
• primary analytical-method or standards documents for measurement claims.

Search snippets, retailer pages, social posts, generic summaries, and AI-generated text are not scientific sources. A review article may help locate a primary paper, but the material claim is checked against the primary or official record before it enters the library.

A source being present does not prove that its methods were adequate, that the result was positive, that the finding was replicated, or that it applies to a Serum Research lot.

Product and approval boundaries

The library does not use a clinical paper as a sales claim. No trial result or FDA approval should appear beside a price, rating, availability badge, product carousel, or purchase control.

A catalog link is optional and separately gated. It appears only after the exact material identity, amount, form, formulation, label, and lot evidence match the entity discussed. If those facts are unresolved, the evidence page remains useful without the link.

The library does not treat these as interchangeable:

• free base and a salt or hydrate;
• native peptide and modified analog;
• DAC and no-DAC forms;
• full-length protein or peptide and a fragment;
• one isoform and another isoform;
• activity units and mass;
• an ingredient and a named finished drug;
• component studies and a multi-component blend; or
• a substance identifier and an approval decision.

What this library does not do

The library does not:

• recommend a compound for a health condition or outcome;
• compare doses, routes, cycles, combinations, or personal protocols;
• provide injection, ingestion, mixing, preparation, reconstitution, or self-experimentation instructions;
• describe a research material as FDA approved because an approved product contains a related ingredient;
• present 99%+, third-party tested, research grade, sterility, or another quality claim without exact lot-and-method support;
• turn animal or laboratory findings into human-benefit promises; or
• hide negative, null, mixed, or missing results to make a record sound more favorable.

Frequently asked questions

Does a ClinicalTrials.gov record prove that a compound works?

No. Registration shows that a study record exists. It does not prove completion, posted results, publication, a favorable outcome, replication, FDA review, or product equivalence.

Does a PMID or DOI mean a study is high quality?

No. These are identifiers. Readers still need to examine the material, model, design, controls, outcomes, effect estimates, precision, missing data, funding, conflicts, and corrections.

Does an FDA UNII mean a substance is approved?

No. A UNII identifies a substance in FDA’s substance-registration system. FDA states that the presence of a UNII does not imply regulatory review or approval.

Can animal evidence be used as proof of a human outcome?

No. Animal work can test a hypothesis in a stated species and model. Translation to people requires direct human evidence and may still be limited by population, intervention, endpoint, and duration.

What happens when a study’s primary result is null?

The null result remains visible. Secondary, exploratory, post-hoc, or subgroup findings are labeled and are not used to overwrite the prespecified primary outcome.

Why might a compound page omit a product link?

Because a shared name is not enough. The catalog material’s identity, form, amount, formulation, label, and lot evidence must support the relationship before a commercial link is added.

How are corrections handled?

Each public record should show a review date and a monitored correction route. A material correction should update the visible page, the source registry, and the change record together rather than silently rewriting the claim.

Corrections and review dates

Scientific and regulatory records change. Dynamic trial and FDA statements are rechecked on publication day and on the cadence specified in the source registry. Stable primary papers are checked for corrections, retractions, expressions of concern, or replaced versions.

Questions about the library or a potential correction can be sent through the Serum Research contact page.

Primary and official sources

1. National Library of Medicine: About PubMed
2. ClinicalTrials.gov: Protocol Registration Data Element Definitions
3. ClinicalTrials.gov: Results Data Element Definitions
4. FDA: About Drugs@FDA
5. FDA: Approved Drugs, Questions and Answers
6. FDA: UNII Search
7. NIH: Enhancing Reproducibility through Rigor and Transparency