How the numbers are counted

Compounds: the peptide profiles in the Peptide Library, one per profile. Peptides are amino-acid chains, modified peptides, peptide hormones and proteins included. Non-peptides in the library have their own pages but are left out of every peptide total.

Publications: PubMed (MEDLINE) records, read through Europe PMC. Preprints and records outside PubMed are not counted. Statistics cover publications first published from January 1, 2000 to the snapshot's cutoff; peptide pages show every year.

Trials: ClinicalTrials.gov registrations that list a peptide as an intervention.

Very large literatures (molecules the body makes in quantity, such as glutathione, growth hormone, NAD+, oxytocin, hCG and IGF-1, and any compound whose search matches more than 15,000 records) are reported as yearly counts only. Most of those papers measure the molecule rather than study it as a compound, so they are never ranked against, or added to, the others. Vendor blends and comparison entries have no literature of their own and are excluded.

Each peptide is searched by its name, synonyms and aliases. Descriptions (“GLP-1/glucagon dual agonist”), combinations, slang shortenings and bare short codes are left out, and names that are also other things need context: p21 is searched only as P021 with neurotrophic wording, because “p21” is also a well-known protein. A term whose own hit count dwarfs every other name of its peptide is held out until an editor reviews it. Every term and the exact queries are in the search-terms dataset, so anyone can rerun them.

A matched record counts for a peptide when its name is in the title, when the peptide is indexed as a MeSH chemical or major topic, or, when the name is only in the abstract, when an AI reading of the abstract found the peptide studied (an intervention, exposure or main subject) with at least 70% confidence. Records that only mention a peptide are counted separately and never in headline figures. Editors can confirm or exclude any match, and their decision always wins. A paper about several peptides counts once in every total and once per peptide in peptide figures.

Every count uses the first publication date (the earlier of e-publication and print), the same as Europe PMC's FIRST_PDATE. The latest year is incomplete: it is always labeled as year to date and compared only with the same months of earlier years, January 1 through the cutoff's month and day. All of PubMed is shown alongside, from whole months, for context.

Each publication has one study type, and each study type belongs to one evidence class, so evidence shares add up to 100% of classified publications. Correction and retraction notices count as records in totals but are left out of every share, as are publications not yet classified. PubMed publication types decide first; MeSH headings decide when no type applies; the MeSH tag “Humans” alone never makes a paper clinical, because it is also applied to work on human cells. What the rules can't decide goes to an AI model (Claude), which reads the record and abstract and answers with a confidence score: answers under 0.7 are not used and go to an editor. Abstracts are used for this only and are never stored or shown. The share of publications classified with AI is published with every dataset.

• GLP-1 receptor agonists

  GLP-1 receptor agonists, including dual and triple agonists (GIP, glucagon) and GLP-1/amylin combinations. Orforglipron is a non-peptide GLP-1 agonist and is shown apart from peptide totals.

  semaglutide, tirzepatide, retatrutide, liraglutide, dulaglutide, exenatide, lixisenatide, efpeglenatide, ecnoglutide, cotadutide, survodutide, mazdutide, pemvidutide, maridebart-cafraglutide, amycretin, cagrisema, vk2735, oxyntomodulin, orforglipron

• Amylin analogues

  Amylin receptor agonists, alone or combined with a GLP-1 agonist.

  cagrilintide, petrelintide, pramlintide, amycretin, cagrisema

• Growth hormone axis

  Growth hormone, long-acting growth hormones, GHRH analogues, ghrelin-receptor agonists (secretagogues), IGF-1 analogues and GH fragments.

  sermorelin, tesamorelin, cjc-1295, cjc-1295-no-dac, ipamorelin, ghrp-1, ghrp-6, pralmorelin-ghrp2, hexarelin, relamorelin, tabimorelin, macimorelin, anamorelin, mk-677, lonapegsomatropin, somapacitan, somatrogon, somatropin, mecasermin, igf-1, igf-1-lr3, igf-1-des, gh-frag-176-191, aod-9604, mgf, peg-mgf

• Tissue-repair peptides

  Peptides studied mainly for wound healing, tissue protection and repair.

  bpc-157, pentadeca-arginate, tb-500, thymosin-beta-4, ghk-cu, ahk-cu, kpv, ara-290, b7-33, mgf, peg-mgf

• Cognitive and neurological peptides

  Compounds the library lists for cognitive or neurological research: neuropeptides, ACTH and tuftsin analogues, neurotrophic peptides and brain-derived peptide preparations.

  semax, n-acetyl-semax-amidate, adamax, selank, n-acetyl-selank-amidate, cerebrolysin, cortexin, dihexa, p21, davunetide, pe-22-28, pacap-38, orexin-a, dsip, pinealon, cortagen, cerluten, cranoluten, humanin, klotho, noopept

• Longevity and mitochondrial peptides

  Compounds researched for aging, senescence and mitochondrial function. Glutathione and NAD+ are counted separately (counts only).

  elamipretide, ss-20, humanin, humanin-g, mots-c, epitalon, foxo4-dri, klotho, gdf-11, atx-304, endoluten, svetinorm, vilon, zhenoluten, pinealon

• Short peptide bioregulators

  Short peptides and peptide complexes from the bioregulator literature (Khavinson and colleagues).

  bonomarlot, bronchogen, cardiogen, cartalax, cerluten, chondromix, chonluten, cortagen, cranoluten, endoluten, epitalon, libidon, livagen, ovagen, pancragen, pinealon, prostamax, suprefort, svetinorm, testagen, testoluten, thyreogen, ventfort, vesugen, vilon, vladonix, zhenoluten

• Thymic peptides

  Thymus-derived peptides and their synthetic analogues.

  thymalin, thymogen, thymolin, thymopentin, thymosin-alpha-1, thymulin, zinc-thymulin

• Melanocortin agonists

  Melanocortin receptor agonists and their fragments.

  afamelanotide, melanotan-ii, bremelanotide, setmelanotide, cosyntropin, kpv

• Cosmetic peptides

  Peptides used as skincare and hair-care ingredients.

  acetyl-hexapeptide-8, acetyl-octapeptide-3, acetyl-tetrapeptide-3, biotinoyl-tripeptide-1, decapeptide-12, egf-oligopeptide-1, myristoyl-pentapeptide-17, palmitoyl-hexapeptide-12, palmitoyl-pentapeptide-4, palmitoyl-tetrapeptide-7, palmitoyl-tripeptide-1, palmitoyl-tripeptide-5, syn-ake, trifluoroacetyl-tripeptide-2

A registration is not a result. A trial counts as reporting results when results are posted on ClinicalTrials.gov by the cutoff or when the registry links a results publication. “Completed trials without published results” covers completed trials whose primary completion was at least 12 months before the cutoff and that have neither. Trial status is as registered on the extraction date.

Records are checked against PubMed's publication types and against Crossref, which includes the Retraction Watch database. Retracted records are left out of every count, and the number left out is published. PMIDs are confirmed against PubMed and DOIs validated with Crossref; disagreements go to review. Two PMIDs sharing one DOI go to review and are never merged automatically.

Every published number comes from a frozen snapshot: the peptides with their exact queries, every matched publication with its classification at that moment, trials and counts. Its version (for example 2026.1) is part of every citation, file and API response. Report statistics are published after an editor approves them. A published snapshot never changes: a correction is a new version that supersedes it, with a revision note, and both stay in the archive. The dashboard's live figures come from an automated weekly snapshot, published only when every automated check passes, and always labeled as automated.

Peptide publications analyzed total_publications

Counted publications (PubMed records about an indexed library peptide: name in the title, indexed as a MeSH chemical or major topic, or confirmed from the abstract; retracted records excluded) first published between the window start and the cutoff, each counted once however many peptides it covers.

Peptides indexed compounds_indexed

Library peptides whose full publication records are indexed (counts-only, excluded and non-peptide compounds left out).

Peptides with at least one counted publication compounds_with_publications

Indexed peptides with one or more counted publications in the window.

Publications so far this year publications_ytd

Counted publications first published from January 1 of the cutoff year through the cutoff date.

Publication growth, year to date growth_ytd

Change in counted publications from the same months of the previous year to this year, January 1 through the cutoff day in both years.Numerator: This year to date minus the same period last yearDenominator: The same period last year

All of PubMed, growth year to date pubmed_growth_ytd

The same comparison for every PubMed record (whole months only), for context: how fast the literature as a whole grew.Numerator: PubMed records this year to date minus the same months last yearDenominator: PubMed records in the same months last year

Average yearly growth over ten years cagr_10y

Compound annual growth rate of counted publications per full calendar year, from ten years before the last full year to the last full year.Numerator: Publications in the last full yearDenominator: Publications ten years earlier

All of PubMed, average yearly growth over ten years pubmed_cagr_10y

The same rate for every PubMed record, for context.

Share of publications that are human clinical research human_clinical_share

Randomized trials, other clinical trials, observational human studies and case reports, as a share of classified counted publications in the window.Numerator: Human clinical publicationsDenominator: Classified publications (all evidence classes except unclassified and notices)

Share of publications that are preclinical preclinical_share

Animal, in vitro and computational studies as a share of classified counted publications in the window.Numerator: Animal + in vitro + computational publicationsDenominator: Classified publications

Share that are animal studies animal_share

Animal studies as a share of classified counted publications in the window.Numerator: Animal studiesDenominator: Classified publications

Share that are in vitro studies in_vitro_share

Cell, tissue and biochemical studies as a share of classified counted publications in the window.Numerator: In vitro studiesDenominator: Classified publications

Share that are reviews and meta-analyses secondary_share

Reviews, systematic reviews and meta-analyses (which report no new primary data) as a share of classified counted publications in the window.Numerator: Reviews, systematic reviews and meta-analysesDenominator: Classified publications

Share not yet classified unclassified_share

Counted publications in the window whose study type neither the rules nor a confident AI answer could decide.Numerator: Unclassified publicationsDenominator: All counted publications in the window

Randomized controlled trial publications rct_publications

Counted publications in the window classified as randomized controlled trials in humans.

Share of peptides with any human clinical publication compounds_with_human_clinical_share

Indexed peptides with at least one counted human clinical publication up to the cutoff, as a share of indexed peptides with any counted publication.Numerator: Peptides with a human clinical publicationDenominator: Peptides with any counted publication

Share of peptides studied only preclinically compounds_preclinical_only_share

Indexed peptides with counted preclinical publications and no human clinical publication up to the cutoff, as a share of indexed peptides with any counted publication.Numerator: Peptides with preclinical but no human clinical publicationsDenominator: Peptides with any counted publication

Registered clinical trials trials_registered

ClinicalTrials.gov registrations first posted by the cutoff that list an indexed peptide as an intervention.

Completed trials without published results trials_without_results_share

Completed trials whose primary completion was at least 12 months before the cutoff that have neither results posted on ClinicalTrials.gov by the cutoff nor a results publication linked in the registry.Numerator: Completed trials without posted or linked resultsDenominator: Completed trials with primary completion 12 or more months before the cutoff

Share of publications that are open access open_access_share

Counted publications in the window that Europe PMC or OpenAlex reports as open access, among those with a known status.Numerator: Open-access publicationsDenominator: Publications with a known open-access status

Retracted publications excluded retracted_publications

Publications about indexed peptides that are retracted (PubMed, Crossref or Retraction Watch) and were left out of every count.

Share classified with AI assistance ai_classified_share

Counted publications in the window whose study type came from a confident AI answer because the PubMed metadata could not decide it.Numerator: Publications with an AI study typeDenominator: All counted publications in the window

Most researched peptides top_compounds

Indexed peptides ranked by counted publications in the window. A publication counts once for each peptide it covers. Volume says nothing about effectiveness or safety.

Most researched research-only peptides top_research_only

The same ranking for peptides with no FDA approval and no trial-stage label in the library, leaving out molecules the body makes itself.

Fastest growing peptides fastest_growing_compounds

Year-to-date growth in counted publications per peptide (same months both years), among peptides with at least the minimum count in the earlier period.

Fastest growing research areas fastest_growing_categories

Year-to-date growth in counted publications per research category (MeSH-based; a paper can be in several), among categories with at least the minimum count in the earlier period.

Publications by research area top_categories

Counted publications in the window per research category. Categories overlap, so they are never added up.

Peptides with the least human evidence evidence_gap

Indexed peptides with at least the minimum number of counted publications, ranked by the lowest share of human clinical publications (all years to the cutoff).

GLP-1 agonist publications glp1_publications

Counted publications in the window about at least one peptide in the GLP-1 receptor agonist group.

GLP-1 agonists, share of all peptide publications glp1_share

Counted publications about GLP-1 group peptides as a share of all counted publications in the window.Numerator: GLP-1 group publicationsDenominator: All counted publications

GLP-1 agonist publication growth, year to date glp1_growth_ytd

Year-to-date growth in counted publications about GLP-1 group peptides.

GLP-1 agonists, share human clinical glp1_human_clinical_share

Human clinical publications among classified counted GLP-1 group publications in the window.Numerator: Human clinical GLP-1 publicationsDenominator: Classified GLP-1 publications

GLP-1 agonist registered trials glp1_trials

Registered trials first posted by the cutoff that list a GLP-1 group peptide as an intervention.

Cognitive and neurological peptide publications cognitive_publications

Counted publications in the window about at least one peptide in the cognitive and neurological group.

Cognitive and neurological peptides, share human clinical cognitive_human_clinical_share

Human clinical publications among classified counted publications about the cognitive and neurological group in the window.Numerator: Human clinical publicationsDenominator: Classified publications in the group

Longevity and mitochondrial peptide publications longevity_publications

Counted publications in the window about at least one peptide in the longevity and mitochondrial group.

Longevity and mitochondrial peptides, share human clinical longevity_human_clinical_share

Human clinical publications among classified counted publications about the longevity group in the window.Numerator: Human clinical publicationsDenominator: Classified publications in the group

Journals publishing the most peptide research top_journals

Journals ranked by counted publications in the window.

Caveat flags: partial_year (year to date), small_baseline (fewer than 30 in the earlier period), high_unclassified (over 10% not yet classified), whole_months_only (PubMed-wide counts).

• More publications, citations or trials do not mean stronger evidence, effectiveness or safety.
• Being studied does not mean being approved; regulatory labels come from the library.
• Preclinical results do not establish safety or effects in people.
• No indexed study does not mean no research exists: the index covers PubMed records that name the peptide.
• Classification is automated and imperfect; editors review flagged records and validation samples are published.

No abstracts or full texts are included in any file.

1. Download the dataset version named in the citation from the data page.
2. The statistics file gives every value with its numerator and denominator; the publications file lists every counted record with its classification and peptides.
3. The open specification (computeStatistics in the public developer kit) recomputes every statistic from the files.
4. The search-terms file gives each peptide's exact queries to rerun against Europe PMC and ClinicalTrials.gov. Counts can move a little as the sources update; a snapshot freezes what was true on its extraction date.