PubMed (NCBI E-utilities)

Returns BioC JSON for one or more PMIDs: the title and abstract as plain text passages, each with the genes, diseases, chemicals, variants, species and cell lines found in it, normalised to NCBI Gene, MeSH, dbSNP and taxonomy identifiers with character offsets.

With no database named, returns the list of every Entrez database. With db set, returns that database's record count, last build date, searchable field tags with their descriptions, and the names of every link it supports.

Resolves a name or fragment to PubTator's normalised entities, returning the @TYPE_Name id, the biotype, the source database and accession (NCBI Gene id, MeSH descriptor, dbSNP rsid, taxonomy id) and which synonym matched.

Searches PubMed by normalised biomedical concept rather than by keyword and returns ranked articles with PMID, PMCID, title, journal, authors, DOI, date, a matching snippet and the entities the paper mentions, together with facet counts over genes, diseases and chemicals.

Returns the complete body text of a PubMed Central open-access article as BioC JSON: every passage tagged with its section type (TITLE, ABSTRACT, INTRO, METHODS, RESULTS, DISCUSS, REF), plus the DOI, the PMID and the licence the text is published under.

Returns the LinkOut records for a PubMed article: every publisher, repository and aggregator that hosts it, each with the resolved URL, the provider name, a categories list such as Full Text Sources or Medical, and an attributes list that says whether it is a free resource or subscription/membership/fee required.

Maps article identifiers onto each other: give it PMIDs, PMCIDs, DOIs or Manuscript IDs and it returns the full identifier set for each one, plus a release date and a status per record.

Returns the linksets Entrez holds for the given UIDs: UIDs in a target database, for example the PMC id of a PubMed article, the PMIDs citing it, the books that cite it, or the gene, protein and clinical-variant records it mentions.

Lists which linksets actually exist for the given PMIDs - the target database, the machine linkname such as pubmed_pmc_refs, the human menu label and a priority - without fetching any of the linked UIDs.

Returns everything NLM records for one MeSH descriptor: its preferred term and every entry-term synonym, the tree numbers that place it in the hierarchy, the scope note defining it, and the qualifiers (subheadings) it may be combined with.

Searches the Medical Subject Headings vocabulary by label and returns the matching descriptors with their canonical D-number URIs.

Returns the entities that a given gene, chemical, disease or variant is linked to by a named relation across all of PubMed, each with the number of publications that support it.

Runs an Entrez query and returns the matching UID list with the total count, the translated query stack and paging offsets - identifiers only, no record content.

Semantic search over individual sentences from PubMed abstracts and PMC full text, returning 100 sentences with a similarity score, their PMID and PMCID, inline entity annotations, and the section they came from (RESULTS, METHODS, DISCUSS - null for abstract and unsectioned text).

Returns the PMIDs PubMed considers related to the given article, each with the numeric similarity score that ranks them - the same computed neighbour set that drives the "Similar articles" panel on pubmed.ncbi.nlm.nih.gov.

Returns the document summary for one or more Entrez UIDs: for PubMed that is title, authors, journal, volume/issue/pages, publication and e-publication dates, article ids including DOI and PMC, publication types and the PubMed status. No abstract text.

Returns the PubMed articles that discuss a given dbSNP variant, with PMID, PMCID, title, journal, authors, date and the matching text snippets from each paper.

Resolves an rsID, gene name or HGVS string to the variants LitVar knows, each with its canonical LitVar id, rsID, gene, HGVS notation, recorded clinical significance and how many PubMed articles mention it.