R/drug_info.R

In diogocamacho/drugminer: DrugMineR: Drug information retrieval tool.

#' Drug info
#' 
#' Given a KEGG query obtained using the \code{\link{kegg_query}} function, returns information on a chemical compound. Returned elements are drug ids for compound (when available), pathways the compound is involved in, chemical formula, and external ids (CAS, PubChem, ChEBI, and DrugBank)  for compound.
#' 
#' @param kegg_content The content of a KEGG query
#' @return A tibble.
#' 
#' @examples 
#' q <- kegg_query(query_string = "aspirin")
#' info <- extract_info(kegg_content = q)
drug_info <- function(drug_content) {
  
  # i'll need to carry this over
  drug_id <- str_extract(string = str_split(string = drug_content, pattern = "NAME")[[1]][1], pattern = "(D\\d{5})")
  
  
  # is it a mixture?
  y1 <- str_split(string = drug_content, pattern = "NAME")[[1]][1]
  y2 <- grep("mixture", y1, ignore.case = TRUE)
  if (length(y2) == 0) {
    drug_mixture <- tibble::tibble(id = drug_id,
                                   mixture = 0)  
  } else {
    drug_mixture <- tibble::tibble(id = drug_id,
                                   mixture = 1)
  }
  
  
  # drug group
  y1 <- str_extract_all(string = drug_content, pattern = "(DG\\d{5})")[[1]]
  if (length(y1) != 0) {
    drug_group <- tibble::tibble(id = drug_id,
                                 drug_group = unique(y1))
  } else {
    drug_group <- tibble::tibble(id = drug_id,
                                 drug_group = NA)
  }
  
  
  # drug efficacy
  y1 <- str_split(str_split(drug_content, "EFFICACY")[[1]][2], "\n")[[1]][1]
  if (!is.na(y1)) {
    drug_efficacy <- str_remove(string = y1, pattern = "^ *")
    drug_efficacy <- unique(tolower(str_remove_all(string = str_split(drug_efficacy, ",")[[1]], pattern = "^ ")))
    drug_efficacy <- tibble::tibble(id = drug_id,
                                    drug_efficacy = drug_efficacy)
  } else {
    drug_efficacy <- tibble::tibble(id = drug_id,
                                    drug_efficacy = NA)
  }
  
  
  # drug targets
  y1 <- str_extract_all(string = drug_content, pattern = "(HSA:\\d{1,})")[[1]]
  y1 <- gsub("HSA:", "", y1)
  drug_targets <- tibble::tibble(id = drug_id,
                                 target_entrez = y1)
  
  # molecular properties
  drug_formula <- str_remove_all(str_split(str_split(drug_content, "FORMULA")[[1]][2], "\n")[[1]][1], " ")
  drug_mass <- as.numeric(str_remove_all(str_split(str_split(drug_content, "EXACT_MASS")[[1]][2], "\n")[[1]][1], " "))
  drug_weight <- as.numeric(str_remove_all(str_split(str_split(drug_content, "MOL_WEIGHT")[[1]][2], "\n")[[1]][1], " "))
  
  mprops <- tibble::tibble(drug_id = drug_id,
                           formula = drug_formula, 
                           exact_mass = drug_mass, 
                           molecular_weight = drug_weight)
  

  # disease indication
  y1 <- str_split(drug_content, "DISEASE")[[1]][2]
  if (!is.na(y1)) {
    drug_disease <- str_extract_all(string = y1, pattern = "(DS:[A-Z]\\d{5})")[[1]]
    drug_disease <- unique(gsub(pattern = "DS:", replacement = "", x = drug_disease))
    drug_disease <- tibble::tibble(id = drug_id,
                                   drug_disease = drug_disease)
  } else {
    drug_disease <- tibble::tibble(id = drug_id,
                                   drug_disease = NA)
  }
  
  # affected pathways
  y1 <- str_extract_all(string = drug_content, pattern = "(hsa\\d{5})")[[1]]
  y1 <- gsub("hsa", "map", y1)
  drug_pathways <- tibble::tibble(id = drug_id,
                                 pathway_id = y1)
  
  # external ids
  cas_id <- str_split(str_split(drug_content, "CAS: ")[[1]][2], "\n")[[1]][1]
  pubchem_sid <- str_split(str_split(drug_content, "PubChem: ")[[1]][2], "\n")[[1]][1]
  chebi_id <- str_split(str_split(drug_content, "ChEBI: ")[[1]][2], "\n")[[1]][1]
  drugbank_id <- str_split(str_split(drug_content, "DrugBank: ")[[1]][2], "\n")[[1]][1]
  ext_ids <- tibble::tibble(drug_id = drug_id,
                            cas_id = cas_id, 
                            pubchem_sid = pubchem_sid,
                            pubchem_cid = sid2cid(pubchem_sid)$pubchem_cid,
                            chebi_id = chebi_id, 
                            drugbank_id = drugbank_id)
  
  return(list(mixture = drug_mixture,
              group = drug_group,
              efficacy = drug_efficacy,
              targets = drug_targets,
              indication = drug_disease,
              pathway = drug_pathways,
              external_ids = ext_ids,
              properties = mprops))
}