Retrieve sample attributes from EBI's ENA repository in JSON format

biosample_attributes.R

library(glue)
library(httr)
library(jsonlite)
library(xml2)

#' Query ENA's REST API for information about records
#' 
#' @param accessions Character vector of one or more ENA record identifiers
#' @return An `xml_document` object
get_records <- function(accessions) {
  r <- httr::GET(
    url = glue::glue(
      "https://www.ebi.ac.uk/ena/browser/api/xml/",
      "{glue::glue_collapse(accessions, sep = ',')}"),
    httr::content_type("text/xml")
  )
  httr::stop_for_status(r)
  httr::content(r, as = "parsed", type = "text/xml",
                encoding = "UTF-8")
}

# return a list of JSON strings with sample attributes
sample_ids <- c("SAMN12207420", "SAMN08358280", "SAMN06470116")

json <- lapply(setNames(sample_ids, sample_ids), \(sample_id) {
  # retrieve the record in XML format
  get_records(sample_id) |> 
    # extract the attributes
    xml2::xml_find_all(xpath = "//SAMPLE_ATTRIBUTES") |>
    xml2::as_list() |> 
    # convert into JSON
    jsonlite::toJSON(auto_unbox = TRUE, pretty = TRUE)
})
json[[2]]  # second sample