Access and Analyze PubMed Biomedical Literature

This skill provides comprehensive guidance for searching PubMed, the U.S. National Library of Medicine's database of MEDLINE and life sciences literature, and accessing it programmatically. It covers advanced search query construction (Boolean operators, field tags like [au]/[ti]/[ab]/[mh]/[pt]/[dp], phrase and proximity searching), MeSH (Medical Subject Headings) controlled-vocabulary searching with subheadings like /therapy and /prevention & control, publication type and date/text-availability filtering, programmatic access via the NCBI E-utilities REST API (ESearch, EFetch, ESummary, EPost, ELink) with example Python request code and documented rate limits (3 req/sec without an API key, 10 req/sec with one), citation matching by PMID/DOI/PMC ID or via ECitMatch, systematic literature review construction using the PICO framework, search history and My NCBI saved-search/alert features, related-article and citation discovery via ELink, and export to reference-manager formats (.nbib, AMA/MLA/APA/NLM styles).

Use it when searching for biomedical or life sciences research articles, constructing complex search queries with Boolean operators, field tags, or MeSH terms, conducting systematic literature reviews or meta-analyses, accessing PubMed data programmatically via the E-utilities API, finding articles by author/journal/date/type, retrieving citation information or abstracts, working with PMIDs or DOIs, or building automated literature-monitoring workflows.

PubMed's database coverage has real limits worth knowing before relying on it: pre-1975 articles often lack abstracts, full author names are only available from 2002 forward, search results display up to 10,000 maximum, search history expires after 8 hours of inactivity, and automatic term mapping can produce unexpected results that should be checked via Search Details. PubMed provides citations and abstracts, not always full text - full-text access depends on the publisher, institutional access, or open-access status. For production API use, obtain and use an API key, respect rate limits, and use the history server for result sets over 500 articles rather than paginating manually.

Input: a research question, disease/topic/author to search, or a PICO-structured systematic review question. Output: a constructed PubMed query string, or programmatic E-utilities API calls returning PMIDs, abstracts, or full records.

base_url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/"
params = {
    "db": "pubmed",
    "term": "diabetes[tiab] AND 2024[dp]",
    "retmax": 100,
    "retmode": "json",
    "api_key": "YOUR_API_KEY"
}
response = requests.get(f"{base_url}esearch.fcgi", params=params)
pmids = response.json()["esearchresult"]["idlist"]

Queries PubMed directly and via the NCBI E-utilities REST API, links to related NCBI databases (GenBank, ClinicalTrials.gov) via ELink, and exports to reference managers like Zotero, Mendeley, and EndNote via .nbib format. Bundles three detailed reference files (api_reference.md, search_syntax.md, common_queries.md) loaded as needed for complex operations.

Researchers, clinicians, and data engineers conducting biomedical literature searches, systematic reviews, or building automated PubMed-based literature monitoring pipelines who need precise query construction and correct E-utilities API usage rather than trial-and-error searching.