Editorial methodology / versioned record
How PSC reads
a research claim.
PSC does not assign a reputation score to a molecule. We ask whether a specific statement is supported by the exact material, model, endpoint, and source offered for it—and we make that reasoning visible.
Exact claim. Exact source. Visible limit.
Each substantive conclusion should identify the material, model or population, endpoint, and supporting citation.
Availability never determines evidence.
Catalog status, price, inventory, and popularity do not change the scientific conclusion or make one material equivalent to another.
Know what kind of answer you are reading.
DOSSIER
A narrow question with a source ledger
Reviewed dossiers separate findings by model, map substantive conclusions to named citations, state limitations beside the finding, display a review date, and publish a change history.
GUIDE
An explanation of how to investigate
Method guides explain molecular identity, analytical methods, study design, endpoints, databases, and regulatory records without turning a research signal into a broader claim.
PASSPORT
A starting point, not a completed review
Discovery passports orient readers to an identity, broad evidence landscape, principal interpretation limit, and neutral public-database searches. They are deliberately not labeled as reviewed dossiers.
Evidence type and evidence strength are not the same thing.
The label identifies the kind of question a source can answer. A rigorous molecular experiment may answer its question better than a weak human report answers another.
IDENTITY
What material is being discussed?
Sequence, structure, chemical form, conjugation, mass, and analytical identity. Identity establishes the object of study—not an organism-level outcome.
EX VIVO
What happened outside an intact organism?
Cell-free, cellular, or tissue-based work. The biological source, assay system, concentration, timing, and controls determine the scope.
MODEL
What happened in a non-human organism?
These findings may generate or refine a hypothesis but cannot be relabeled as a human result.
HUMAN
What can a constrained observation answer?
Small, uncontrolled, retrospective, abstract-only, formulation-specific, or otherwise limited records retain those limitations beside the finding.
HUMAN
What did a defined protocol measure?
Prospective research with a defined population, comparator, endpoint, and protocol. Conclusions remain restricted to that protocol.
RECORD
What did a regulator review—and where?
A live, jurisdiction-, product-, formulation-, route-, indication-, and date-specific record. Evidence for a reviewed product does not establish equivalence for a separate research vial.
OR MIXTURE
Can the name resolve to one exact material?
Used when a name is ambiguous, the record is sparse, or a listing contains multiple components. Component evidence is not evidence for an exact mixture.
Primary and official sources first.
- Live regulator records and official scientific databases.
- Peer-reviewed original reports for the exact material and endpoint.
- Registered protocols and trial records where available.
- Systematic reviews used to map a field and expose limitations.
- Representative mechanistic or preclinical studies used only for model-specific statements.
Supplier pages, promotional summaries, social posts, testimonials, and unattributed AI output are not used as evidence for a reviewed conclusion. Review articles are assessed by the design and quality of the underlying sources, not by the word “review.”
PubMed is a literature-retrieval resource; indexing is not an approval or quality verdict. ClinicalTrials.gov is a sponsor-submitted registry; registration does not validate a hypothesis and may contain no results. PubChem supports chemical identity and cross-references; it cannot verify a commercial vial.
About PubMed ↗ · How to read a trial record ↗ · About PubChem ↗
A close name is not proof of equivalence.
Before connecting a publication to a listing, PSC checks the available name against sequence or structure, chemical form, terminal modifications, conjugates, counter-ions, formulation, and mixture composition. When those fields are unresolved, the uncertainty remains visible.
Four questions before a sentence is ready.
What exact material, sequence, structure, chemical form, or preparation was studied?
Was the work molecular, in vitro, ex vivo, animal, observational human, controlled human, or regulatory?
What was measured, when, against what comparator, and with what uncertainty?
Which exact record supports the statement, and what does that source not support?
Where the record cannot answer one of these questions, PSC does not fill the gap with a stronger-sounding conclusion.
Corrections clarify the record; they do not disappear it.
Reviewed pages display a last-reviewed date. Time-sensitive regulatory and registry references are checked during scheduled reviews and when a material source changes.
The date of change, the statement that changed, why it changed, and the source supporting the revision. Editorial simplification may occur without changing the scientific meaning.
Coverage metadata updated. The published-dossier count now reflects all six reviewed dossiers available in the research library.
Method version 1.0 published. Distinguished reviewed dossiers, method guides, and discovery passports; defined evidence labels, identity matching, source hierarchy, public corrections, and commercial conflict controls.
PSC sells research materials. That relationship is disclosed.
Operating a commercial catalog creates a potential conflict. Catalog availability, price, popularity, inventory, or sales activity does not determine topic selection, evidence labeling, or the conclusion of a reviewed dossier.
Educational content does not establish the identity, purity, safety, efficacy, approval, outcome, or intended use of a catalog item. Evidence for an active ingredient, published preparation, or branded product does not transfer automatically to a separate research material.
Discovery is useful only when its limits are visible.
Each passport provides an identity check, broad research orientation, principal interpretation limit, and neutral searches in public databases. It does not claim an exhaustive search, complete source screening, formal risk-of-bias assessment, current regulatory determination, or full synthesis of every publication.
If a statement cannot preserve the distinction between identity, mechanism, experimental model, human record, and regulatory status, it is not ready to publish as a reviewed conclusion.