data-raw/drug_annot/pug_view/parse_pug.R

In hms-dbmi/drugseqr.data: Data for 'dseqr' and 'rkal' Packages

#' Check Pubchem PUG-View for GRAS notice
#'
#' @param pug_view List from reading PUG-View JSON data with \code{\link[rjson]{fromJSON}}
#'
#' @return Boolean indicating if \code{pug_view} has a GRAS notice.
#' @keywords internal
#'
#'
check_gras <- function(pug_view) {

    # default
    is_gras <- FALSE

    # get top level headings
    sections <- pug_view$Record$Section
    headings <- sapply(sections, `[[`, 'TOCHeading')

    # check for index of food heading
    foodi <- which(headings == 'Food Additives and Ingredients')

    if (length(foodi)) {
        food_sections <- sections[[foodi]]$Section
        food_headings <- sapply(food_sections, `[[`, 'TOCHeading')
        is_gras <- 'FDA Generally Recognized as Safe - GRAS Notices' %in% food_headings
    }

    return(is_gras)
}

#' Extract DrugBank URL from PUG-View
#'
#' @inheritParams check_gras
#'
#' @return Character vector with DrugBank ID or NA if none exists
#' @keywords internal
get_drugbank <- function(pug_view) {

    # default
    db_id <- NA_character_

    # get reference source names
    refs <- pug_view$Record$Reference
    sources <- sapply(refs, `[[`, 'SourceName')

    # get first drugbank index
    dbi <- which(sources == 'DrugBank')[1]

    if (!is.na(dbi)) {
        # get base url
        db_url <- refs[[dbi]]$URL
        db_id <- gsub('^[^#]+/(DB[0-9]+)', '\\1', db_url)
    }
    return(db_id)
}


#' Extract Wikipedia URL from PUG-View
#'
#' @inheritParams check_gras
#'
#' @return Character vector with Wikipedia page
#' @keywords internal
get_wikipedia <- function(pug_view) {

    # default
    wiki_name <- NA_character_

    # get reference source names
    refs <- pug_view$Record$Reference
    sources <- sapply(refs, `[[`, 'SourceName')

    # get first drugbank index
    wiki <- which(sources == 'Wikipedia')[1]

    if (!is.na(wiki)) {
        # get base url
        wiki_url <- refs[[wiki]]$URL

        # make sure is wikipedia (wikidata like Pubchem but worse)
        if (grepl('https://en.wikipedia.org/wiki', wiki_url, fixed = TRUE))
            wiki_name <- gsub('^https://en.wikipedia.org/wiki/([^/]+)$', '\\1', wiki_url)
    }
    return(wiki_name)
}