R/ttd_check.R

In neurogenomics/MultiEWCE: Multi-Scale Target Explorer

#' Gene therapy targets
#'
#' Identify the overlap between your prioritised list of gene therapy targets
#' and currently existing gene therapy targets that are currently on the market
#' or are in clinical trials. Uses data from the
#' \href{https://db.idrblab.net/ttd/full-data-download}{
#' Therapuetic Target Database}.
#' @param top_targets Top targets generated by
#'  \link[MSTExplorer]{prioritise_targets}.
#' @param drug_types Filter results by drug type.
#' @param keep_status Filter results by drug approval status.
#' @param failed_status Drug approval status categories that indicate
#' the drug failed.
#' @param remove_status Remove results by drug approval status.
#' @inheritParams plot_
#' @inheritParams load_example_ctd
#' @inheritParams data.table::merge.data.table
#'
#' @export
#' @importFrom utils tail
#' @examples
#' top_targets <- MSTExplorer::example_targets$top_targets
#' res <- ttd_check(top_targets=top_targets)
ttd_check <- function(top_targets,
                      drug_types = NULL,
                      failed_status = c(
                        "Terminated",
                        "Withdrawn from market",
                        NA
                      ),
                      keep_status = NULL,
                      remove_status=c(NA),
                      allow.cartesian = FALSE,
                      show_plot = TRUE,
                      save_path = NULL,
                      height=NULL,
                      width=NULL,
                      phenotype_to_genes=HPOExplorer::load_phenotype_to_genes()){
  # top_targets <- prioritise_targets()$top_targets
  # drug_types <- c("Gene therapy"
    # "Antisense drug",
    # "Antisense oligonucleotide",
    # "Aptamer",
    # "Combination drug (Antisense drug)",
    # "CAR T Cell Therapy",
    # "CAR T Cell Therapy (Dual specific)",
    # "CAR-NK Cell therapy",
    # "CAR-NKT Cell therapy",
    # "CAR-PBMC Cell therapy",
    # "mRNA therapy",
    # "RNAi therapeutics",
    # "Short hairpin RNA",
    # "siRNA drug",
    # "Small activating RNA",
    # "Small interfering RNA",
    # "TCR-T cell therapy",
    # "Peptide",
    # "Protein",
    # "Protein/peptide",
    # "Protein/peptide drug"
  # )

  TARGETID <- DRUGNAME <- DRUGTYPE <- DRUGID <-
    GENENAME2 <- prioritised <- HIGHEST_STATUS <- NULL;

  ttdi <- KGExplorer::get_ttd()
  #### Remove results that can't be linked to specific genes #####
  dat_sub <- ttdi$merged[!is.na(TARGETID) &
                         !is.na(GENENAME2) &
                         GENENAME2!="",]
  #### Filter by drug type ####
  if(!is.null(drug_types)){
    dat_sub <- dat_sub[
      grepl(paste(drug_types,collapse = "|"),DRUGNAME,ignore.case = TRUE) |
      grepl(paste(drug_types,collapse = "|"),DRUGTYPE,ignore.case = TRUE),]
  }
  #### Filter by status ####
  if(!is.null(keep_status)){
    dat_sub <- dat_sub[HIGHEST_STATUS %in% keep_status,]
  }
  if(!is.null(remove_status)){
    dat_sub <- dat_sub[!HIGHEST_STATUS %in% remove_status,]
  }
  dat_sub[,failed:=HIGHEST_STATUS %in% failed_status]
  #### Filter to only those in top_targets ####
  dat_sub2 <- (merge(
        dat_sub[failed==FALSE],
        top_targets,
        allow.cartesian = allow.cartesian,
        by.x = "GENENAME3",
        by.y = "gene_symbol")[,c("GENENAME2","TARGETID","TARGNAME",
                                 "INDICATI","DRUGID","DRUGNAME",
                                 "HIGHEST_STATUS",
                                 "disease_name","disease_id","hpo_name",
                                 "CellType","ontLvl")] |>
    unique())
  #### Count proportion of drugs that our analyses captured ####
  pct_captured <- length(unique(dat_sub2$DRUGID)) /
    length(unique(dat_sub$DRUGID))*100
  # length(unique(paste0(dat_sub2$DRUGID,dat_sub2$INDICATI,
  #                      dat_sub2$GENENAME2)))
  dat_sub[,prioritised:=(DRUGID %in% dat_sub2$DRUGID)]
  #### Hypergeometric test ####
  dat_sub[,failed:=HIGHEST_STATUS %in% failed_status]
  fail <- dat_sub[failed==TRUE,drop=FALSE]
  notfail <- dat_sub[failed==FALSE,drop=FALSE]
  ttd_hypergeo_out <- ttd_hypergeo(fail=fail,
                                   notfail=notfail,
                                   top_targets=top_targets,
                                   p2g=phenotype_to_genes)
  #### Plot ####
  plt <- plot_ttd(dat_sub = dat_sub,
                  failed_status = failed_status)
  #### Show ####
  if(isTRUE(show_plot)) methods::show(plt)
  #### Save ####
  KGExplorer::plot_save(plt = plt,
                        save_path=save_path,
                        height=height,
                        width=width)
  #### Return ####
  return(
    list(data=dat_sub,
         data_overlap=dat_sub2,
         pct_captured,
         plot=plt,
         ttd_hypergeo_out=ttd_hypergeo_out)
  )
}