We use privacy-friendly analytics. Read our privacy policy.
Research library · sources first
Browse papers, human trials, reviews, observational research, and mechanistic studies with direct links to PubMed, DOI records, or original sources. Grades and library-wide metrics are kept as context; the research itself comes first.
Source-first browsing
They are grouped by study design so stronger clinical evidence is easy to distinguish from observational, mechanistic, and preclinical evidence. Supportive, mixed, null, and unfavorable findings are all eligible.
Evidence synthesis
Start here when you want research that evaluates multiple studies together. Review quality still depends on the studies included, but these sources help show the broader evidence picture.
L-theanine
Background evidence
DOI · 10.1093/nutrit/nuaf054 ↗Curr Med Chem · Curcumin + piperine
Background evidence
PubMed · PMID 38561618 ↗Avian Pathol · Artemisia Absinthium
Background evidence
PubMed · PMID 38616734 ↗Explore (NY) · Ashwagandha Withania somnifera
Background evidence
PubMed · PMID 39348746 ↗Complementary Therapies in Medicine · Chamomile
Background evidence
Original source ↗Phytotherapy Research · Asian Ginseng
Background evidence
PubMed · PMID 39474788 ↗Human intervention research
Direct links to intervention studies in people. These are especially useful for checking the tested population, preparation, dose, duration, comparator, and measured outcome.
Osteoarthritis Cartilage · Boswellia
Background evidence
PubMed · PMID 40554037 ↗Nutrients · Mulberry Leaf
Background evidence
PubMed · PMID 38892603 ↗Nutrients · Guayusa
Background evidence
PubMed · PMID 38140277 ↗J Med Food · Fucoxanthin
Background evidence
PubMed · PMID 37405785 ↗Gut Microbes · Psyllium
Background evidence
PubMed · PMID 37078654 ↗Drug Discov Ther · Andrographis Paniculata
Background evidence
PubMed · PMID 37821383 ↗Human context
Human evidence outside controlled trials can reveal associations, real-world patterns, and clinical signals. It can be informative without establishing causation on its own.
Moringa
A 2026 narrative review mapped 22 clinical trials and 9 published case reports. Human findings were heterogeneous and preparation-specific; the review describes generally good tolerability but notes rare hyp…
Background evidence
PubMed · PMID 41710586 ↗Noni
Published case reports describe acute hepatitis temporally associated with noni products. The signal is low-incidence/uncertain-causality but clinically relevant enough to retain a liver caution.
Background evidence
PubMed · PMID 16094725 ↗How it might work
Animal, in-vitro, and mechanistic studies can explain plausible pathways and generate hypotheses. They belong in the evidence map, but they are intentionally kept separate from proof of a human benefit.
Magnesium glycinate
Background evidence
DOI · 10.3390/nu17050748 ↗Cells · bacopaside ii
Background evidence
PubMed · PMID 30037060 ↗Passionflower
The extract showed activity at adenosine A1/A2A and 5-HT2A-related targets and inhibited monoamine uptake in vitro.
Background evidence
PubMed · PMID 42462921 ↗Longan
A 2025 study reported sleep-related enzyme/pathway effects in cell models using longan flower ethanol extract. It is not human sleep evidence and is plant-part specific, so it should be labeled mechanistic/p…
Background evidence
PubMed · PMID 40722663 ↗Where the evidence is strongest
These are current indexable ingredient profiles graded Strong (A) or Moderate (B). A grade is a navigation aid, not a guarantee of benefit, and different outcomes or preparations can have different evidence.
Research tools
Study-level source trail
Search structured study records and follow the evidence back to the ingredient pages and source identifiers that use them.
Open →Ingredient-level evidence
Look up an ingredient and inspect its current evidence grade, source coverage, safety context, and research maturity.
Open →Visual research map
Explore how botanicals, compounds, pathways, and evidence relationships connect across the research library.
Open →Aggregate analysis
See the library-wide metrics, evidence-grade distribution, ambiguity analysis, and downloadable public dataset.
Open →Independent research databases
Good research should be easy to check elsewhere. These external databases are useful for finding primary papers, trial registrations, reviews, and government evidence summaries.
Library context · dataset v2026.08.16
922
deduplicated structured study records
282
human-evidence source records
922
study records with a direct PubMed, DOI, or source URL