Methodology
Data source
Papers are collected from the PubMed / Europe PMC public bibliographic databases. Search queries are built primarily onEurope PMC's MeSH (Medical Subject Headings) controlled vocabulary, supplemented with title-text search to catch newly published papers whose MeSH indexing has not yet been applied.
Search queries are not narrowed by the direction of the result. The goal is to collect as much of the relevant literature as practical, not to find papers that support a particular conclusion.
What is machine-derived and what is left to a human
| Item | Derivation |
|---|---|
| Subject (human / animal) | Machine (MeSH Humans / Animals check tags) |
| Study design | Machine (publication type + MeSH priority) |
| In vitro | Human-reviewed only — no corresponding MeSH check tag exists in the sample, confirmed mechanically |
| Topic relevance | Machine (weighted match on chemical registry number, MeSH major topic, title, abstract) |
| Collection domain (3-way) | Machine (mapped from topic) |
| Inclusion (study-design criteria) | A policy written by a person, applied mechanically — no individual paper is hand-picked |
Relevance scoring
Appearing in a search result does not by itself establish topical relevance, so matches are weighted by where they occur. A matching chemical registry number (UNII) or a MeSH major-topic designation are judgments the source itself made about what the paper is about, and each is sufficient on its own. A title-only or abstract-only match is not sufficient alone and requires corroboration.
Coverage
| Category | Count |
|---|---|
| Papers indexed (total) | 8,488 |
| Adverse-event research | 2,281 |
| Research on formulated ingredients | 4,477 |
| Research on non-formulated ingredients | 1,827 |
| Abstract shown (CC BY licensed) | 2,280 |
| Abstract withheld — link out only | 6,208 |
Counts reflect only what actually exists in this page's data (public/data). This is not the same thing as a search engine's raw hit count for the underlying population.
About this section
This site is built on the principle of assembling substance rather than writing narrative. Listing papers here is not intended as a claim of any effect, and the papers collected do not themselves prove efficacy. Adverse-event literature is collected as an independent body of evidence, not as a point of comparison.
This site is a literature database. It does not sell or endorse any product. Papers are listed based on bibliographic metadata from their respective sources; this site does not evaluate or warrant the findings of any individual study. Consult a medical professional for health decisions.