Map PubTator3 Annotations to Abstract Sentences — pubtator

Splits abstract text into sentences and assigns each PubTator3 entity annotation to its containing sentence via character-offset overlap. When available, the PubTator3 passage text and offsets are used directly.

pubtator_sentences(pubtations)

Arguments

pubtations: A data.table returned by get_records(endpoint = "pubtations").

Value

A data.table with annotation columns plus integer sentence_id, sentence, sentence_start, and sentence_end. sentence_start and sentence_end are zero-based, end-exclusive entity offsets within sentence. PubTator passage metadata columns are used for mapping but are not returned. Only passage annotations that can be assigned to a sentence are returned.

Examples

if (FALSE) { # \dontrun{
pmids <- search_pubmed('"Biomarkers Consortium"')

pubtations <- get_records(pmids, endpoint = "pubtations")

mapped <- pubtator_sentences(pubtations)
} # }