Getting BibTeX bibliography from PubMed with R, rentrez, and glue

pubmed2bibtex.ipynb

rentrez2bibtex.r

## Getting BibTeX bibliography from PubMed with R, rentrez, and glue
library(rentrez)
library(tidyverse)
library(magrittr)
library(glue)

## search term
search_term = "\"flegontov p\"[au] OR \"flegontov pn\"[au]"

## number of publications
npubs = entrez_search(db = "pubmed", term = search_term, retmax = 100)$count
## get ncbi uids
ncbi_ids = entrez_search(db = "pubmed", term = search_term, retmax = npubs)$ids #%>% str
## get all publications as json, save into list object
esummary = entrez_summary(db = "pubmed", id = ncbi_ids, retmode = "json")

## extract items into table
papers <- esummary %>% 
    lapply(extract, c("uid", "title", "fulljournalname", "source","pubdate", "sortpubdate", "issue", "volume", "pages")) %>% 
    bind_rows %>% 
    mutate(
        journalname = str_remove(fulljournalname, ":.*"),
        year = str_extract(sortpubdate, "\\d{4}"),
        pages = str_replace(pages, "-","--")
    )

papers$doi <- esummary %>% 
    lapply(function(x) use_series(x, articleids) %>% subset(idtype == "doi", select = value) %>% pluck(1))

papers$authors <- esummary %>% 
    lapply(function(x) use_series(x, authors) %>% pull(name) %>% paste(collapse = ", "))

#papers %>% glimpse

bibtex <- papers %>% 
    glue_data("
@article{{{uid},
    author={{{authors}}},
    year={{{year}}},
    title={{{title}}},
    journal={{{source}}},
    number={{{issue}}},
    volume={{{volume}}},
    pages={{{pages}}},
    doi={{{doi}}}
}}
")

## write the bibtex string into a file
writeLines(bibtex, "papers.bib")

## remove empty fields
system("sed -i '/{}/d' papers.bib")

rentrez2bibtex.yml

name: pubmed2bibtex
channels:
 - conda-forge
dependencies:
 - r-base=4
 - r-tidyverse # includes magrittr & glue
 - r-rentrez
## pick an IDE
# - rstudio-desktop
# - jupyterlab
# - r-irkernel