How to Search PubMed for Peptide Research: A Practical Guide

In plain English

PubMed is a free search engine for biomedical research, run by the US National Library of Medicine. It holds more than 40 million citations and abstracts. It does not hold the full papers itself, but it often links to them ^[1].

Typing a few words into the search box works, but PubMed quietly rewrites the search. It adds related terms and index terms called MeSH (Medical Subject Headings, a fixed list of topic labels) ^{[2, 5]}. That can help or hurt, so it is worth checking what PubMed actually searched.

Good searches have three habits. They use every name for a substance, including codes and spellings. They combine labels and free-text words. And they are written down, so someone else can run the same search later ^{[10, 12]}.

The short version

• PubMed is a citation database, not a full-text library ^[1].
• Check “Search Details” to see how a query was translated ^[2].
• Search MeSH and title/abstract words together ^{[10, 11]}.
• Check every key paper for retractions and corrections ^[6].

What PubMed contains

Two consequences follow. First, a PubMed record is not proof that a journal is MEDLINE-indexed: in a 2019 analysis, 91% of PubMed content was indexed in MEDLINE, and the share had slowly fallen as PMC grew ^[14]. Second, records that are not indexed for MEDLINE, such as preprints and online-ahead-of-print citations, lack MeSH terms, so filters that depend on MeSH can miss them ^[2].

How PubMed reads a search: Automatic Term Mapping

When words are typed without tags, PubMed tries to match them, in order, against a subject translation table (which includes MeSH), a journals translation table, the author index and an investigator index. Once a term matches, mapping stops ^[2]. The subject table holds British and American spellings, singular and plural forms, synonyms, MeSH entry terms and other related terms ^[2].

Example. On 25 October 2026, the untagged search trifluoroacetate returned 7,143 records. PubMed’s translation included "trifluoroacetic acid"[MeSH Terms], "trifluoroacetates"[All Fields] and "trifluoroacetic"[All Fields], among others. The translation is shown under Search Details on the Advanced Search page ^[2].

Automatic Term Mapping is switched off by quotation marks and by wildcards ^[2]. That is useful for precision, but it means a quoted or truncated term no longer picks up MeSH.

MeSH terms and subheadings

MeSH is a controlled, hierarchically organised vocabulary produced by the National Library of Medicine for indexing, cataloguing and searching biomedical information ^[5]. It is updated every year ^[2]. MEDLINE articles are now indexed with MeSH by an automated algorithm ^[2].

• Explosion. A MeSH search automatically includes the more specific terms beneath it in the hierarchy. Adding :noexp turns this off ^[2]. On 25 October 2026, "Peptides"[mh] returned 3,021,895 records and "Peptides"[mh:noexp] returned 190,867.
• Major topic. [majr] limits retrieval to records where the term is one of the main topics of the article ^[2].
• Subheadings. A subheading narrows a MeSH term to one aspect, written as Term/Subheading. "Peptides/chemical synthesis"[mh] returned 33,295 records; the same pair as a major topic returned 14,303 ^[2].

Do not rely on MeSH alone. In a study of 73 systematic reviews, limiting both Embase and MEDLINE searches to major thesaurus terms reduced the number of references by 13%, but 15% of the reviews lost more than 5% of their relevant references; limiting both databases to title and abstract words lost more than 5% in 33% of reviews ^[11]. A published method for building searches works by comparing what the thesaurus terms retrieve with what the free-text words retrieve, to find missing candidate terms ^[10].

Field tags worth knowing

All of these appear in the PubMed User Guide’s list of search fields ^[2]. For example, Page MJ[au] AND PRISMA[ti] AND 2021[dp] returned 13 records on 25 October 2026.

Boolean operators, phrases, truncation and proximity

• Boolean operators must be typed in capitals: AND, OR, NOT ^[2]. Use brackets to group synonyms: (A OR B) AND C.
• Phrases go in double quotes. If the phrase is not in PubMed’s phrase index, the quotes are ignored, the words are processed by Automatic Term Mapping, and a “Quoted phrase not found in phrase index” message appears ^[2].
• Truncation uses an asterisk, which stands for zero or more characters. A term must have at least four characters before the first wildcard ^[2]. On 25 October 2026, tri*[tiab] was searched simply as "tri"[Title/Abstract]: the wildcard was dropped.
• Proximity finds words within N words of each other, in any order, in the title, title/abstract or affiliation fields: "search terms"[tiab:~N] ^[2]. "peptide adsorption"[tiab] returned 274 records; "peptide adsorption"[tiab:~2] returned 464.

Worked examples

All counts below were recorded on 25 October 2026. They will grow, because citations are added to PubMed seven days a week ^[3].

Example 1: trifluoroacetate counter-ions in peptides

An untagged search, trifluoroacetate peptide counterion, returned 30 records. A structured version, with each concept written out in full, returned 42:

(trifluoroacetate*[tiab] OR "trifluoroacetic acid"[tiab] OR "Trifluoroacetic Acid"[mh])
AND (peptide*[tiab] OR "Peptides"[mh])
AND (counterion*[tiab] OR counter-ion*[tiab] OR "salt form*"[tiab] OR "salt exchange"[tiab])

Adding AND 2015:2026[dp] cut this to 18. Search Details showed that PubMed rewrote counter-ion* as "counter ion*", handling the hyphenated term as a phrase, as the User Guide describes for hyphenated wildcards ^[2]. For the chemistry itself, see Counter-Ions in Peptides.

Example 2: peptide adsorption to labware

(adsorption[tiab] OR adsorptive[tiab] OR "non-specific binding"[tiab] OR "nonspecific binding"[tiab])
AND peptide*[tiab]
AND (polypropylene[tiab] OR "glass vial*"[tiab] OR labware[tiab] OR "sample container*"[tiab])

This returned 39 records. Relevant papers may use other words for the same effect, so the next step is to read the first relevant hits, note the terms their authors use, and add them as synonyms ^[10]. Background: Peptide Adsorption to Labware.

Example 3: every name for one compound

Using angiotensin II, a well-characterised reference peptide:

Neither the MeSH term nor the text word alone found everything. For a research peptide, the list of names should include the International Nonproprietary Name, development codes, abbreviations, salt forms and spelling variants; see How Peptides Are Named and CAS Numbers and Chemical Identifiers.

Filters, sorting, “Similar articles” and citations

• Article type filters mostly rely on the Publication Type field. The Systematic Review filter is an exception: it also uses a search strategy, and can be added as AND systematic[sb] ^[2].
• Filters can exclude relevant records. Many filters depend on MeSH assigned during MEDLINE indexing, so citations that are not yet indexed may be left out ^[2]. The “Other animals” species filter, for example, is the search "animals"[mh:noexp] ^[2].
• Best Match is the default sort order ^[2]. It is a machine-learning ranking trained on past user searches; its most important signals are past usage of an article, publication date, relevance score and article type ^[9]. Sorting changes the order of records, not which records are found.
• Similar articles are generated by comparing words from the title, abstract and MeSH terms, and filters are not applied to them ^[2].
• Cited by lists other publications that cite a record, using data submitted by publishers ^[2].
• Search History on the Advanced page can be downloaded as a CSV file; it expires after 8 hours of inactivity ^[2].

PMID, PMCID and DOI: three different identifiers

A PMCID shows that full text is archived in PMC. It does not by itself show that the text was peer reviewed, because PMC also contains preprints ^[4].

Checking for retractions and corrections

NLM links each article’s citation to citations for its retraction notices, errata, expressions of concern and comments ^[6].

NLM classes corrections, corrigenda, addenda and partial retractions all as errata ^[6]. On 25 October 2026, "Retracted Publication"[pt] returned 35,220 records and "Expression of Concern"[pt] returned 4,056.

Retraction does not stop citation. In a study of 7,813 retracted papers indexed in PubMed, only 722 of 13,252 citation contexts written after the retraction (5.4%) acknowledged it ^[15]. Open the PubMed record of every key paper before relying on it. Journals and retraction databases are covered in Peer Review, Retractions and Predatory Journals.

Making a search repeatable

PRISMA-S, an extension of the PRISMA reporting guideline, sets out 16 reporting items so that each part of a literature search is completely reported and therefore reproducible ^[12]. The PRESS 2015 checklist for peer-reviewing a search covers six elements: translation of the research question; Boolean and proximity operators; subject headings; text-word searching; spelling, syntax and line numbers; and limits and filters ^[13]. A practical record includes:

1. The database and interface (PubMed web or the API).
2. The exact search string, copied, not retyped.
3. The date the search was run and the number of records.
4. Every filter and limit applied ^[12].

The E-utilities API. The E-utilities are the public API to NCBI’s Entrez system, covering PubMed and PMC among other databases. They are a suite of eight server-side programs that accept a fixed URL syntax for search, link and retrieval ^[7]. A scripted search makes the record automatic:

https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=pubmed&retmode=json&term=%22peptide+adsorption%22%5Btiab%5D

ESearch returns the record count, a list of PMIDs and PubMed’s translation of the query; ESummary and EFetch then return citation details and abstracts for those PMIDs. NCBI publishes usage guidelines for the E-utilities, which scripted searches should follow ^[7]. Pooling what a search finds is covered in Systematic Reviews and Meta-Analyses.

What this means in Australia

• Free access. Searching PubMed is free of charge and does not require registration, so it is open to any Australian researcher or reader ^[3].
• A listing is not an approval. A PubMed record shows that a paper exists. It says nothing about whether a substance may be supplied in Australia. The TGA states that unapproved peptide products have not been assessed by the TGA for safety, quality or effectiveness, and that a “research use only” disclaimer does not make supply lawful ^[16].
• Check the register too. Products on the Australian Register of Therapeutic Goods can be looked up in the TGA’s ARTG search ^[17]; see What Is the ARTG?.

This is general information, not legal advice. Laws change; check the current Poisons Standard and TGA guidance.

Frequently asked questions

Is PubMed free to use?

Yes. Searching PubMed is free and needs no registration. PubMed holds citations and abstracts; full text is often available through PMC or the publisher ^{[1, 3]}.

What is the difference between PubMed and MEDLINE?

MEDLINE is the largest component of PubMed: journals selected by NLM and indexed with MeSH. PubMed also includes PMC and Bookshelf records, some of which are not MEDLINE-indexed ^{[1, 14]}.

What are MeSH terms?

MeSH (Medical Subject Headings) is NLM’s controlled, hierarchical vocabulary used to index articles. Searching a MeSH term also searches the more specific terms beneath it unless [mh:noexp] is used ^{[2, 5]}.

Why do PubMed record counts change over time?

New citations are added seven days a week, and indexing and query translation can change, so the same string can return different counts on different dates. Recording the date and count makes this visible ^{[3, 12]}.

How can a reader tell if a paper has been retracted?

Open its PubMed record and look for a “Retraction in:” link or the Retracted Publication publication type. Corrections appear as “Erratum in:” and expressions of concern have their own link ^[6].

What is the difference between a PMID, a PMCID and a DOI?

A PMID identifies a PubMed record, a PMCID identifies a full-text article in PMC, and a DOI is a persistent identifier that resolves through doi.org. PMC’s ID Converter translates between them ^{[7, 8]}.

References

1. National Center for Biotechnology Information. About PubMed. US National Library of Medicine. pubmed.ncbi.nlm.nih.gov/about. Accessed 25 October 2026.
2. National Center for Biotechnology Information. PubMed User Guide. US National Library of Medicine. pubmed.ncbi.nlm.nih.gov/help. Accessed 25 October 2026.
3. US National Library of Medicine. About MEDLINE. nlm.nih.gov. Accessed 25 October 2026.
4. National Center for Biotechnology Information. About PMC. US National Library of Medicine. pmc.ncbi.nlm.nih.gov. Accessed 25 October 2026.
5. US National Library of Medicine. Welcome to Medical Subject Headings (MeSH). nlm.nih.gov. Accessed 25 October 2026.
6. US National Library of Medicine. Errata, retractions, and other linked citations in PubMed. nlm.nih.gov. Accessed 25 October 2026.
7. National Center for Biotechnology Information. APIs: Entrez Programming Utilities (E-utilities) and PMC APIs. ncbi.nlm.nih.gov. Accessed 25 October 2026.
8. DOI Foundation. What is a DOI? doi.org. Accessed 25 October 2026.
9. Fiorini N, Canese K, Starchenko G, et al. Best Match: new relevance search for PubMed. PLoS Biol. 2018;16(8):e2005343. doi:10.1371/journal.pbio.2005343. PMID: 30153250
10. Bramer WM, de Jonge GB, Rethlefsen ML, Mast F, Kleijnen J. A systematic approach to searching: an efficient and complete method to develop literature searches. J Med Libr Assoc. 2018;106(4):531–541. doi:10.5195/jmla.2018.283. PMID: 30271302
11. Bramer WM, Giustini D, Kleijnen J, Franco OH. Searching Embase and MEDLINE by using only major descriptors or title and abstract fields: a prospective exploratory study. Syst Rev. 2018;7(1):200. doi:10.1186/s13643-018-0864-9. PMID: 30458825
12. Rethlefsen ML, Kirtley S, Waffenschmidt S, et al. PRISMA-S: an extension to the PRISMA Statement for Reporting Literature Searches in Systematic Reviews. Syst Rev. 2021;10(1):39. doi:10.1186/s13643-020-01542-z. PMID: 33499930
13. McGowan J, Sampson M, Salzwedel DM, et al. PRESS Peer Review of Electronic Search Strategies: 2015 guideline statement. J Clin Epidemiol. 2016;75:40–46. doi:10.1016/j.jclinepi.2016.01.021. PMID: 27005575
14. Ossom Williamson P, Minter CIJ. Exploring PubMed as a reliable resource for scholarly communications services. J Med Libr Assoc. 2019;107(1):16–29. doi:10.5195/jmla.2019.433. PMID: 30598645
15. Hsiao TK, Schneider J. Continued use of retracted papers: temporal trends in citations and (lack of) awareness of retractions shown in citation contexts in biomedicine. Quant Sci Stud. 2022;2(4):1144–1169. doi:10.1162/qss_a_00155. PMID: 36186715
16. Therapeutic Goods Administration. Understanding your responsibilities when importing, compounding and supplying unapproved peptide products. Safety advisory, 13 April 2026. tga.gov.au. Accessed 25 October 2026.
17. Therapeutic Goods Administration. Australian Register of Therapeutic Goods (ARTG) search. tga.gov.au/resources/artg. Accessed 25 October 2026.