R/pharm_oncox.R
In sigven/oncoPharmaDB: Molecularly targeted cancer drugs and biomarkers
Defines functions
get_biomarkers
get_drugs
Documented in
get_biomarkers
get_drugs
#' Get antineoplastic drugs and drug regimens
#'
#' @description
#' Downloads preprocessed datasets to a local cache directory and returns a
#' selected set of drugs based on various criteria set by the user.
#'
#' The dataset comes as a `list` object, with two elements:
#'
#' * `metadata` - a data frame with metadata regarding drug resources used
#' * `records` - a data frame with drug records
#'
#' @param cache_dir local cache directory for data retrieval
#' @param force_download force download data from remote repository even if
#' data exists in cache
#' @param exclude_salt_forms exclude salt forms of drugs
#' @param exclude_adc exclude antibody-drug conjugates (ADCs)
#' @param treatment_category main treatment category, classified according to
#' ATC or not ('targeted_therapy_classified',
#' 'targeted_therapy_unclassified','chemo_therapy_classified','hormone_therapy_classified',
#' 'immuno_suppressants_classified','other')
#' @param drug_is_approved logical indicating if resulting drug records
#' should contain approved drugs only
#' @param drug_target character vector with drug targets (gene symbols)
#' for drug records included in results
#' @param drug_action_type character vector with drug action types to
#' include in drug record list -
#' possible values "INHIBITOR","AGONIST","MODULATOR","ANTAGONIST",
#' "BLOCKER","ACTIVATOR","BINDING AGENT","OPENER",
#' "STABILISER","CROSS-LINKING AGENT",DISRUPTING AGENT","OTHER"
#' @param drug_action_inhibition logical indicating to only return drug records
#' with inhibitory mechanism-of-action
#' @param drug_indication_main character vector with main tumor types for
#' which drug(s) are indicated. Possible values: "Adrenal Gland","Biliary Tract",
#' "Bladder/Urinary Tract","Bone","Breast","Cervix","CNS/Brain",
#' "Colon/Rectum","Esophagus/Stomach","Eye","Head and Neck","Kidney",
#' "Liver","Lung","Lymphoid","Myeloid","Ovary/Fallopian Tube",
#' "Pancreas","Penis","Peripheral Nervous System","Peritoneum",
#' "Pleura","Prostate","Skin","Soft Tissue","Testis","Thymus",
#' "Thyroid","Uterus","Vulva/Vagina"
#' @param drug_source_opentargets logical indicating if resulting drug records
#' should contain drug records from Open Targets Platform/ChEMBL only
#' @param drug_cancer_indication logical indicating if resulting drug records
#' should be for those indicated for cancer conditions only (from approved
#' conditions, found in clinical trials etc.)
#' @param drug_classified_cancer logical indicating if resulting drug
#' records should be for those classified only in the "L" class of ATC (
#' "ANTINEOPLASTIC AND IMMUNOMODULATING AGENTS") only
#' @param drug_has_blackbox_warning logical indicating if resulting drug
#' records should contain drugs with black box warnings only
#' @param drug_approval_year only include records for drugs approved
#' later than this date (year)
#' @param drug_minimum_phase_any_indication only include drug records that
#' are in a clinical phase (any indication) greater or equal than this phase
#' @param output_resolution dictate output record resolution
#' ('drug','drug2target','drug2target2indication')
#'
#' @return
#'
#' The `records` data frame contains the following columns (only selected columns
#' will be shown based on the value of `output_resolution`)
#'
#' \itemize{
#' \item \emph{drug_id} - drug identifier (pharmaOncoX)
#' \item \emph{drug_name} - primary drug name (upper case, NCI Thesaurus)
#' \item \emph{drug_type} - type of drug molecule (Antibody, small molecule etc)
#' \item \emph{molecule_chembl_id} - ChEMBL compound identifier
#' \item \emph{drug_action_type} - main action elicited by drug
#' (antagonist, inhibitor, stabiliser etc)
#' \item \emph{drug_alias} - collection of unambiguous drug aliases
#' (separated by '|')
#' \item \emph{nci_concept_definition} - detailed description of drug
#' mechanism-of-action (NCI Thesaurus)
#' \item \emph{opentargets} - logical - drug is found in the Open Targets
#' Platform resource
#' \item \emph{is_salt} - logical - drug record represents a salt form
#' (excluded by default)
#' \item \emph{is_adc} - logical - drug record represents an
#' antibody-drug conjugate (ADC - excluded by default)
#' \item \emph{drug_blacbox_warning} - logical indicating if drug has
#' blackbox warning
#' \item \emph{nci_t} - NCI thesaurus identifier
#' \item \emph{target_symbol} - gene symbol of drug target
#' \item \emph{target_entrezgene} - Entrez gene identifier of drug target
#' \item \emph{target_genename} - gene name/description of drug target
#' \item \emph{target_ensembl_gene_id} - Ensembl gene identifier of drug target
#' \item \emph{target_type} - type of drug target (single protein, protein
#' family etc.)
#' \item \emph{drug_max_phase_indication} - maximum clinical phase for drug
#' (given indication)
#' \item \emph{drug_approved_indication} - logical indicating if drug has
#' an approved indication
#' \item \emph{drug_frac_cancer_indications} - fraction of drug indications that
#' are for cancers
#' \item \emph{drug_approved_noncancer} - logical indicating if drug
#' is approved for a non-cancer disease
#' \item \emph{drug_n_indications} - number of indications for the
#' given drug (from approved indications, clinical trials etc)
#' \item \emph{drug_year_first_approval} - year drug was first approved
#' \item \emph{drug_max_ct_phase} - maximum clinical phase for drug
#' (any indication)
#' \item \emph{disease_efo_id} - EFO (Experimental Factor Ontology)
#' identifier for drug indication
#' \item \emph{disease_efo_label} - EFO (Experimental Factor Ontology)
#' label for drug indication
#' \item \emph{primary_site} - primary tumor site/type (obtained through
#' https://github.com/sigven/oncoPhenoMap)
#' \item \emph{drug_clinical_id} - drug clinical identifier
#' (clinicaltrials.gov, DailyMed, FDA etc.)
#' \item \emph{drug_clinical_source} - underlying source for drug entry
#' (DailyMed, clinicaltrials.gov, FDA etc.)
#' \item \emph{atc_code_level1} - drug identifier ATC (level 1)
#' \item \emph{atc_level1} - drug label ATC (level 1)
#' \item \emph{atc_code_level1} - drug identifier ATC (level 2)
#' \item \emph{atc_level2} - drug label ATC (level 2)
#' \item \emph{atc_code_level3} - drug identifier ATC (level 3)
#' \item \emph{atc_level3} - drug label ATC (level 3)
#' \item \emph{atc_treatment_category} - treatment category (targeted/chemo/hormone, cancer/other etc)
#' }
#'
#'
#'
#' @export
#'
get_drugs <- function(
cache_dir = NA,
force_download = FALSE,
exclude_salt_forms = TRUE,
exclude_adc = FALSE,
treatment_category =
c("targeted_therapy_classified",
"targeted_therapy_unclassified",
"chemo_therapy_classified",
"hormone_therapy_classified",
"immuno_suppressants_classified",
"other"),
drug_is_approved = FALSE,
drug_target = NULL,
drug_action_type = NULL,
drug_indication_main = NULL,
drug_source_opentargets = FALSE,
drug_cancer_indication = TRUE,
drug_classified_cancer = TRUE,
drug_has_blackbox_warning = FALSE,
drug_approval_year = 1939,
drug_minimum_phase_any_indication = 0,
output_resolution = "drug2target2indication",
drug_action_inhibition = F){
## set logging layout
lgr::lgr$appenders$console$set_layout(
lgr::LayoutFormat$new(timestamp_fmt = "%Y-%m-%d %T"))
valid_output_resolutions <-
c("drug",
"drug2target",
"drug2target2indication")
valid_treatment_categories <-
c("hormone_therapy_classified",
"targeted_therapy_classified",
"targeted_therapy_unclassified",
"chemo_therapy_classified",
"immuno_suppressants_classified",
"other")
is_valid_treatment_category <-
unique(treatment_category %in% valid_treatment_categories)
valid_drug_action_types <-
c("INHIBITOR",
"AGONIST",
"MODULATOR",
"ANTAGONIST",
"BLOCKER",
"ACTIVATOR",
"BINDING AGENT",
"OPENER",
"STABILISER",
"CROSS-LINKING AGENT",
"DISRUPTING AGENT",
"OTHER")
valid_indications_main <-
c("Adrenal Gland",
"Biliary Tract",
"Bladder/Urinary Tract",
"Bone",
"Breast",
"Cervix",
"CNS/Brain",
"Colon/Rectum",
"Esophagus/Stomach",
"Eye",
"Head and Neck",
"Kidney",
"Liver",
"Lung",
"Lymphoid",
"Myeloid",
"Ovary/Fallopian Tube",
"Pancreas",
"Penis",
"Peripheral Nervous System",
"Peritoneum",
"Pleura",
"Prostate",
"Skin",
"Soft Tissue",
"Testis",
"Thymus",
"Thyroid",
"Uterus",
"Vulva/Vagina")
arg_validation_messages <- list()
arg_validation_messages[[1]] <-
assertthat::validate_that(
is.numeric(drug_approval_year),
msg = "ERROR: Argument 'drug_approval_year' must be of type 'numeric'")
arg_validation_messages[[2]] <-
assertthat::validate_that(
drug_approval_year >= 1939 & drug_approval_year <= 2024,
msg = "ERROR: Argument 'drug_approval_year' must be larger than 1939 and less than or equal to 2024")
arg_validation_messages[[3]] <-
assertthat::validate_that(
is.logical(drug_is_approved),
msg = "ERROR: Argument 'drug_is_approved' must be of type 'logical'")
arg_validation_messages[[4]] <-
assertthat::validate_that(
is.logical(drug_source_opentargets),
msg = "ERROR: Argument 'drug_source_opentargets' must be of type 'logical'")
arg_validation_messages[[5]] <-
assertthat::validate_that(
is.logical(drug_classified_cancer),
msg = "ERROR: Argument 'drug_classified_cancer' must be of type 'logical'")
arg_validation_messages[[6]] <-
assertthat::validate_that(
is.logical(drug_cancer_indication),
msg = "ERROR: Argument 'drug_cancer_indication' must be of type 'logical'")
arg_validation_messages[[7]] <-
assertthat::validate_that(
is.character(treatment_category),
msg = "ERROR: Argument 'treatment_category' must be of type 'character'")
arg_validation_messages[[8]] <-
assertthat::validate_that(
length(is_valid_treatment_category) == 1 &
is_valid_treatment_category == T,
msg = paste0("ERROR: Argument 'treatment_category' must contain any ",
"combination of 'chemo_therapy','hormone_therapy', or 'targeted_therapy'"))
arg_validation_messages[[9]] <-
assertthat::validate_that(
is.logical(drug_has_blackbox_warning),
msg = "ERROR: Argument 'drug_has_blackbox_warning' must be of type 'logical'")
arg_validation_messages[[10]] <-
assertthat::validate_that(
output_resolution %in% valid_output_resolutions,
msg = "ERROR: Argument 'output_resolution' must be either 'drug','drug2target', or 'drug2target2indication'")
arg_counter <- 11
if (!is.null(drug_action_type)) {
arg_counter <- arg_counter + 1
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
is.character(drug_action_type),
msg = "ERROR: Argument 'drug_action_type' must be a vector of type 'character'")
arg_counter <- arg_counter + 1
if (length(unique(drug_action_type %in% valid_drug_action_types)) > 1) {
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
F,
msg = paste0(
"ERROR: Argument 'drug_action_type' must be a character vector",
" with any of the following action types: ",
paste(valid_drug_action_types, collapse = ", ")))
arg_counter <- arg_counter + 1
}
}
if (!is.null(drug_indication_main)) {
#arg_counter <- arg_counter + 1
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
is.character(drug_indication_main),
msg = "ERROR: Argument 'drug_indication_main' must be a vector of type 'character'")
arg_counter <- arg_counter + 1
if (length(unique(drug_indication_main %in% valid_indications_main)) > 1) {
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
F,
msg = paste0(
"ERROR: Argument 'drug_indication_main' must be a character vector",
" with any of the following action types: ",
paste(valid_indications_main, collapse = ", ")))
arg_counter <- arg_counter + 1
}
}
if (!is.null(drug_target)) {
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
is.character(drug_target),
msg = "ERROR: Argument 'drug_target' must be a vector of type 'character'")
arg_counter <- arg_counter + 1
}
i <- 1
error_messages <- c()
while (i <= length(arg_validation_messages)) {
if (!is.logical(arg_validation_messages[[i]])) {
error_messages <- c(error_messages, arg_validation_messages[[i]])
}
i <- i + 1
}
if (length(error_messages) > 0) {
lgr::lgr$info('\n')
lgr::lgr$info(error_messages, sep = "\n")
return()
}
## Get full list of drug records
all_drug_recs <- get_drug_records(cache_dir, force_download)
drug_records <- as.data.frame(all_drug_recs[['records']])
metadata <- as.data.frame(all_drug_recs[['metadata']])
if (exclude_salt_forms == TRUE) {
drug_records <- drug_records |>
dplyr::filter(.data$is_salt == FALSE)
}
if (exclude_adc == TRUE) {
drug_records <- drug_records |>
dplyr::filter(.data$is_adc == FALSE)
}
## targeted + chemo + hormone
treatment_category_regex <- ""
atc_treatment_strings <- c()
if("hormone_therapy_classified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer_hormone_therapy")
}
if("chemo_therapy_classified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer_chemo_therapy")
}
if("targeted_therapy_classified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer(_adc)?_targeted_therapy")
}
if("targeted_therapy_unclassified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer_unclassified")
}
if("immuno_suppressants_classified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer_immuno_suppressants")
}
if("other" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "(other_targeted_therapy|unknown)")
}
treatment_category_regex <- paste0(
"^(", paste(atc_treatment_strings, collapse="|"), ")$")
#cat(paste0('Treatment category regex: ', treatment_category_regex))
#cat('\n')
drug_records <- drug_records |>
dplyr::filter(
!is.na(.data$atc_treatment_category) &
stringr::str_detect(
.data$atc_treatment_category,
treatment_category_regex))
if (drug_source_opentargets == TRUE) {
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(.data$opentargets == TRUE)
}
}
if (drug_cancer_indication == TRUE){
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(stringr::str_detect(
.data$drug_cancer_relevance, "by_cancer_"
))
}
}
if (drug_classified_cancer == TRUE){
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(
!is.na(.data$atc_treatment_category) &
stringr::str_detect(
.data$atc_treatment_category, "cancer_"
)
)
}
}
if (drug_action_inhibition == TRUE) {
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(.data$inhibition_moa == TRUE)
}
}
if (drug_has_blackbox_warning == TRUE) {
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(is.na(.data$drug_blackbox_warning) |
.data$drug_blackbox_warning == TRUE)
if (nrow(drug_records) == 0) {
lgr::lgr$info(
paste0("WARNING: For the conditions listed below, NO drugs were found with a blackbox warning\n"))
lgr::lgr$info(
paste0("Condition 1: Drug treatment category: ",
paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(
paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info('\n')
}
}
}
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(is.na(.data$drug_year_first_approval) |
.data$drug_year_first_approval >= drug_approval_year)
if (nrow(drug_records) == 0) {
lgr::lgr$info(
paste0("WARNING: For the conditions listed below, NO drugs were found with an approval date greater than or equal to: ",
drug_approval_year),"\n")
lgr::lgr$info(
paste0("Condition 1: Drug treatment category: ",
paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(
paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info(
paste0("Condition 3: Drugs with blackbox warnings only: ", drug_has_blackbox_warning, "\n"))
lgr::lgr$info('\n')
}
}
if (drug_minimum_phase_any_indication > 0) {
if (nrow(drug_records) > 0) {
## include indications customly retrieved in DailyMed (assuming they are all max phase)
drug_records <- drug_records |>
dplyr::filter(
(!is.na(.data$drug_max_ct_phase) &
.data$drug_max_ct_phase >= drug_minimum_phase_any_indication))
if (nrow(drug_records) == 0) {
lgr::lgr$info(paste0(
"WARNING: For the conditions listed below, NO drugs were found",
"with a clinical phase greater or equal than: ",
drug_minimum_phase_any_indication),"\n")
lgr::lgr$info(
paste0("Condition 1: Drug treatment category: ",
paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(
paste0("Condition 2: Open Targets Platform only: ",
drug_source_opentargets, "\n"))
lgr::lgr$info(
paste0("Condition 3: Drugs with blackbox warnings only: ",
drug_has_blackbox_warning, "\n"))
lgr::lgr$info(
paste0("Condition 4: Drugs with approval later than only: ",
drug_approval_year, "\n"))
lgr::lgr$info('\n')
}
}
}
if (drug_is_approved == TRUE) {
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(!is.na(.data$drug_approved_indication) &
.data$drug_approved_indication == TRUE)
}
if (nrow(drug_records) == 0) {
lgr::lgr$info(
paste0("WARNING: For the conditions listed below, NO approved drugs were found\n"))
lgr::lgr$info(
paste0("Condition 1: Drug treatment category: ",
paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(
paste0("Condition 2: Open Targets Platform only: ",
drug_source_opentargets, "\n"))
lgr::lgr$info(
paste0("Condition 3: Drugs with blackbox warnings only: ",
drug_has_blackbox_warning, "\n"))
lgr::lgr$info(
paste0("Condition 4: Drugs with approval later than only: ",
drug_approval_year, "\n"))
lgr::lgr$info(
paste0("Condition 5: Drugs with minimum clinical phase (any indication) only: ",
drug_minimum_phase_any_indication, "\n"))
}
}
all_drug_targets <- data.frame()
if (!is.null(drug_target)) {
for (target in drug_target) {
df <- data.frame('target_symbol' = target,
stringsAsFactors = F)
all_drug_targets <- dplyr::bind_rows(
all_drug_targets, df
)
}
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(!is.na(.data$target_symbol)) |>
dplyr::inner_join(
all_drug_targets, by = "target_symbol")
if (nrow(drug_records) == 0) {
lgr::lgr$info(paste0("WARNING: For the conditions listed below, NO molecularly targeted drugs were found for the target proteins provided in the 'drug_target' argument: ",
paste(all_drug_targets$target_symbol, collapse = ", "),"\n"))
lgr::lgr$info(paste0("Condition 1: Drug treatment category: ", paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info(paste0("Condition 3: Drugs with approved indications only: ", drug_is_approved, "\n"))
lgr::lgr$info(paste0("Condition 4: Drugs with blackbox warnings only: ", drug_has_blackbox_warning, "\n"))
lgr::lgr$info(paste0("Condition 5: Drugs with approval later than only: ", drug_approval_year, "\n"))
lgr::lgr$info(paste0("Condition 6: Drugs with minimum clinical phase (any indication) only: ", drug_minimum_phase_any_indication, "\n"))
lgr::lgr$info('\n')
}else{
drug_records <- drug_records |>
dplyr::arrange(
dplyr::desc(.data$drug_max_ct_phase),
dplyr::desc(.data$drug_year_first_approval))
}
}
}
if (!is.null(drug_action_type)) {
if (nrow(drug_records) > 0) {
drug_records_action_type <- data.frame()
drug_records$atype <- drug_records$drug_action_type
for (actype in drug_action_type) {
recs <- drug_records |>
dplyr::filter(
!is.na(.data$atype) &
stringr::str_detect(.data$atype, actype)
)
if (nrow(recs) > 0) {
drug_records_action_type <- drug_records_action_type |>
dplyr::bind_rows(recs)
}
}
if (nrow(drug_records_action_type) == 0) {
lgr::lgr$info(paste0("WARNING: For the conditions listed below, no drug records were found for the action types provided in the argument 'drug_action_type': "
, paste(drug_action_type, collapse = ", ")),"\n")
lgr::lgr$info(paste0("Condition 1: Drug treatment category: ", paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info(paste0("Condition 3: Drugs with approved indications only: ", drug_is_approved, "\n"))
lgr::lgr$info(paste0("Condition 4: Drugs with blackbox warnings only: ", drug_has_blackbox_warning, "\n"))
lgr::lgr$info(paste0("Condition 5: Drugs with approval later than: ", drug_approval_year, "\n"))
lgr::lgr$info(paste0("Condition 6: Drugs with minimum clinical phase (any indication) only: ", drug_minimum_phase_any_indication, "\n"))
if (nrow(all_drug_targets) > 0) {
lgr::lgr$info(paste0("Condition 7: Drugs targeted towards the following targets: ", paste(all_drug_targets$target_symbol, collapse = ", "), "\n"))
}
lgr::lgr$info('\n')
}else{
drug_records <- drug_records_action_type |>
dplyr::select(-.data$atype)
}
}
}
if (!is.null(drug_indication_main)) {
if (nrow(drug_records) > 0) {
drug_records_indication <- data.frame()
for (ind in drug_indication_main) {
recs <- drug_records |>
dplyr::filter(
!is.na(.data$primary_site) &
stringr::str_detect(.data$primary_site, ind)
)
if (nrow(recs) > 0) {
drug_records_indication <- drug_records_indication |>
dplyr::bind_rows(recs)
}
}
if (nrow(drug_records_indication) == 0) {
lgr::lgr$info(paste0("WARNING: For the conditions listed below, no drug records were found for the indications (main tumor types) provided in the argument 'drug_indication_main': "
, paste(drug_indication_main, collapse = ", ")),"\n")
lgr::lgr$info(paste0("Condition 1: Drug treatment category: ", paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info(paste0("Condition 3: Drugs with approved indications only: ", drug_is_approved, "\n"))
lgr::lgr$info(paste0("Condition 4: Drugs with blackbox warnings only: ", drug_has_blackbox_warning, "\n"))
lgr::lgr$info(paste0("Condition 5: Drugs with approval later than: ", drug_approval_year, "\n"))
lgr::lgr$info(paste0("Condition 6: Drugs with minimum clinical phase (any indication) only: ", drug_minimum_phase_any_indication, "\n"))
if (nrow(all_drug_targets) > 0) {
lgr::lgr$info(paste0("Condition 7: Drugs targeted towards the following targets only: ", paste(all_drug_targets$target_symbol, collapse = ", "), "\n"))
}
if (!is.null(drug_action_type)) {
lgr::lgr$info(paste0(
"Condition 8: Drugs with the following action types only: ",
paste(drug_action_type, collapse = ", "), "\n"))
}
lgr::lgr$info('\n')
}else{
drug_records <- drug_records_indication
}
}
}
lgr::lgr$info(
paste0("Record set satisfying user-defined criteria: n = ",
nrow(drug_records)))
if (output_resolution == "drug") {
lgr::lgr$info(
"Collapsing record set - providing output on a 'per_drug' resolution")
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::select(-c(.data$target_symbol,
.data$target_ensembl_gene_id,
.data$target_entrezgene,
.data$target_type,
.data$target_genename,
.data$disease_efo_id,
.data$cui,
.data$cui_name,
.data$drug_approved_indication,
.data$drug_clinical_source)) |>
dplyr::distinct()
drug_records <- as.data.frame(
drug_records |>
dplyr::group_by(
dplyr::across(-c(
.data$atc_code_level1,
.data$atc_level1,
.data$atc_code_level2,
.data$atc_level2,
.data$atc_code_level3,
.data$atc_level3,
.data$atc_treatment_category,
.data$disease_efo_label,
.data$primary_site,
.data$drug_clinical_id,
.data$drug_max_phase_indication))) |>
dplyr::summarise(
drug_clinical_id = paste(
unique(sort(.data$drug_clinical_id)), collapse = "|"),
disease_indication = paste(
unique(sort(.data$disease_efo_label)), collapse = "|"),
disease_indication_max_phase = paste(
sort(unique(.data$drug_max_phase_indication),
decreasing = T), collapse = "|"),
disease_main_group = paste(
unique(sort(.data$primary_site)), collapse = "|"),
atc_code_level1 = paste(
unique(sort(.data$atc_code_level1)),
collapse = "|"
),
atc_level1 = paste(
unique(sort(.data$atc_level1)),
collapse = "|"
),
atc_code_level2 = paste(
unique(sort(.data$atc_code_level2)),
collapse = "|"
),
atc_level2 = paste(
unique(sort(.data$atc_level2)),
collapse = "|"
),
atc_code_level3 = paste(
unique(sort(.data$atc_code_level3)),
collapse = "|"
),
atc_level3 = paste(
unique(sort(.data$atc_level3)),
collapse = "|"
),
atc_treatment_category = paste(
unique(sort(.data$atc_treatment_category)),
collapse = "|"
),
.groups = "drop") |>
dplyr::distinct()
)
}
lgr::lgr$info(
paste0("Final record set: n = ", nrow(drug_records), " records"))
}
if (output_resolution == "drug2target") {
lgr::lgr$info("Collapsing record set - providing output on a 'per_drug_target' resolution")
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::select(-c(.data$disease_efo_id,
.data$cui,
.data$cui_name,
.data$drug_approved_indication,
.data$drug_clinical_source)) |>
dplyr::distinct()
drug_records <- as.data.frame(
drug_records |>
dplyr::group_by(
dplyr::across(
-c(.data$disease_efo_label,
.data$atc_code_level1,
.data$atc_level1,
.data$atc_code_level2,
.data$atc_level2,
.data$atc_code_level3,
.data$atc_level3,
.data$atc_treatment_category,
.data$primary_site,
.data$drug_clinical_id,
.data$drug_max_phase_indication))) |>
dplyr::summarise(
drug_clinical_id = paste(
unique(sort(.data$drug_clinical_id)), collapse = "|"),
disease_indication = paste(
unique(sort(.data$disease_efo_label)), collapse = "|"),
disease_indication_max_phase = paste(
sort(unique(.data$drug_max_phase_indication),
decreasing = F), collapse = "|"),
disease_main_group = paste(
unique(sort(.data$primary_site)), collapse = "|"),
atc_code_level1 = paste(
unique(sort(.data$atc_code_level1)),
collapse = "|"
),
atc_level1 = paste(
unique(sort(.data$atc_level1)),
collapse = "|"
),
atc_code_level2 = paste(
unique(sort(.data$atc_code_level2)),
collapse = "|"
),
atc_level2 = paste(
unique(sort(.data$atc_level2)),
collapse = "|"
),
atc_code_level3 = paste(
unique(sort(.data$atc_code_level3)),
collapse = "|"
),
atc_level3 = paste(
unique(sort(.data$atc_level3)),
collapse = "|"
),
atc_treatment_category = paste(
unique(sort(.data$atc_treatment_category)),
collapse = "|"
),
.groups = "drop") |>
dplyr::distinct()
)
}
lgr::lgr$info(
paste0("Final record set: n = ", nrow(drug_records), " records"))
}
if (output_resolution == "drug2target2indication") {
lgr::lgr$info(
"Collapsing record set - providing output on a 'per_drug_target_indication' resolution")
if (nrow(drug_records) > 0) {
drug_records <- as.data.frame(
drug_records |>
dplyr::select(-c(.data$cui, .data$cui_name)) |>
dplyr::distinct() |>
dplyr::group_by(
dplyr::across(
-c(.data$drug_clinical_id,
.data$atc_code_level1,
.data$atc_level1,
.data$atc_code_level2,
.data$atc_level2,
.data$atc_code_level3,
.data$atc_level3,
.data$atc_treatment_category,
.data$drug_clinical_source))) |>
dplyr::summarise(
drug_clinical_id = paste(
unique(sort(.data$drug_clinical_id)), collapse = "|"),
drug_clinical_source = paste(
unique(sort(.data$drug_clinical_source)), collapse = "|"),
atc_code_level1 = paste(
unique(sort(.data$atc_code_level1)),
collapse = "|"
),
atc_level1 = paste(
unique(sort(.data$atc_level1)),
collapse = "|"
),
atc_code_level2 = paste(
unique(sort(.data$atc_code_level2)),
collapse = "|"
),
atc_level2 = paste(
unique(sort(.data$atc_level2)),
collapse = "|"
),
atc_code_level3 = paste(
unique(sort(.data$atc_code_level3)),
collapse = "|"
),
atc_level3 = paste(
unique(sort(.data$atc_level3)),
collapse = "|"
),
atc_treatment_category = paste(
unique(sort(.data$atc_treatment_category)),
collapse = "|"
),
.groups = "drop") |>
dplyr::distinct()
)
}
lgr::lgr$info(
paste0("Final record set: n = ", nrow(drug_records), " records"))
}
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::select(
.data$drug_id,
.data$drug_name,
.data$drug_type,
.data$molecule_chembl_id,
.data$drug_action_type,
.data$drug_alias,
.data$nci_concept_definition,
.data$opentargets,
dplyr::everything()
) |>
dplyr::arrange(
dplyr::desc(.data$drug_max_ct_phase),
dplyr::desc(.data$drug_year_first_approval),
dplyr::desc(.data$opentargets),
nchar(.data$drug_name))
if(!is.null(drug_indication_main)){
if(output_resolution != "drug2target2indication"){
drug_records <- drug_records |>
dplyr::arrange(
dplyr::desc(.data$disease_indication_max_phase),
dplyr::desc(.data$drug_year_first_approval),
dplyr::desc(.data$opentargets),
nchar(.data$drug_name))
}
}
}
oncodrugs <- list()
oncodrugs[['records']] <- drug_records
oncodrugs[['metadata']] <- metadata
return(oncodrugs)
}
#' Get curated cancer biomarker datasets
#'
#' @description
#' Downloads preprocessed datasets to a local cache directory and returns a
#' curated set of genomic biomarkers from multiple sources
#' (CIViC, CGI, MitelmanDB)
#'
#' The dataset comes as a `list` object, with three elements:
#'
#' * `metadata` - a data frame with metadata regarding drug resources used
#' * `data` - a list with four elements ('civic','cgi','mitelmandb','custom_fusions')
#' * `fpath` - path to cache file
#'
#' @param cache_dir Local directory for data download
#' @param force_download Logical indicating if local cache should force downloaded
#' (i.e. set to TRUE to re-download even if data exists in cache)
#'
#' @return
#'
#' Each entry of the source-specific (e.g. 'civic') entry in the `data` list contains
#' a list of three data frames:
#'
#' \itemize{
#' \item \emph{variant} - list of all biomarker variants, extensively populated
#' according to variant aliases (identifer - column \strong{variant_id})
#' \item \emph{clinical} - cross-references between variants recorded in
#' the `variant` data frame and clinical evidence items (identifier -
#' column \strong{evidence_id}) and underlying literature evidence
#' (identifier - column \strong{source_id})
#' \item \emph{literature} - lists literature for all source_id's listed in
#' the `clinical` data frame
#' }
#'
#' @export
#'
get_biomarkers <- function(cache_dir = NA,
force_download = F) {
lgr::lgr$appenders$console$set_layout(
lgr::LayoutFormat$new(timestamp_fmt = "%Y-%m-%d %T"))
if (is.na(cache_dir)) {
lgr::lgr$fatal(paste0("Argument cache_dir = '",
cache_dir, "' is not defined"))
stop()
}
if (!dir.exists(cache_dir)) {
lgr::lgr$fatal(paste0("Argument cache_dir = '",
cache_dir, "' does not exist"))
stop()
}
biomarker_datasets <- list()
file_maps <- c('biomarkers')
for (elem in file_maps) {
fname_local <- file.path(
cache_dir,
paste0(elem,"_v",
db_id_ref[db_id_ref$name == elem,]$pVersion,
'.rds')
)
fname_gd <- googledrive::as_id(
db_id_ref[db_id_ref$name == elem,]$gid)
md5checksum_package <-
db_id_ref[db_id_ref$name == elem,]$md5Checksum
#dat <- NULL
if (file.exists(fname_local) & force_download == F) {
biomarker_datasets[[elem]] <- readRDS(fname_local)
biomarker_datasets[[elem]][['fpath']] <- fname_local
if (!is.null(biomarker_datasets[[elem]][['data']]) &
!is.null(biomarker_datasets[[elem]][['metadata']])) {
lgr::lgr$info(paste0(
"Reading from cache_dir = '",
cache_dir, "', argument force_download = F"))
lgr::lgr$info(paste0("Object '",elem,"' sucessfully loaded"))
}
}else{
googledrive::drive_deauth()
lgr::lgr$info("Downloading remote dataset from Google Drive to cache_dir")
dl <- googledrive::with_drive_quiet(
googledrive::drive_download(
fname_gd,
path = fname_local,
overwrite = TRUE)
)
md5checksum_remote <- dl$drive_resource[[1]]$md5Checksum
md5checksum_local <- tools::md5sum(fname_local)
names(md5checksum_local) <- NULL
if (md5checksum_remote == md5checksum_local) {
biomarker_datasets[[elem]] <- readRDS(fname_local)
biomarker_datasets[[elem]]$fpath <- fname_local
if (!is.null(biomarker_datasets[[elem]][['data']]) &
!is.null(biomarker_datasets[[elem]][['metadata']])) {
lgr::lgr$info(paste0(
"Reading from cache_dir = ' (",
cache_dir, "'), argument force_download = F"))
lgr::lgr$info(paste0("Object '", elem, "' sucessfully loaded"))
lgr::lgr$info(paste0("md5 checksum is valid: ", md5checksum_remote))
}
}else{
lgr::lgr$error(paste0("md5 checksum of local file (", md5checksum_local,
") is inconsistent with remote file (",
md5checksum_remote,")"))
stop()
}
}
}
return(biomarker_datasets[['biomarkers']])
}
#' Tidy eval helpers
#'
#' <https://cran.r-project.org/web/packages/dplyr/vignettes/programming.html>
#'
#' @name tidyeval
#' @keywords internal
#' @importFrom rlang .data :=
NULL
utils::globalVariables(c("."))
sigven/oncoPharmaDB documentation built on Dec. 24, 2024, 9:46 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions get_biomarkers get_drugs
Documented in get_biomarkers get_drugs
#' Get antineoplastic drugs and drug regimens
#'
#' @description
#' Downloads preprocessed datasets to a local cache directory and returns a
#' selected set of drugs based on various criteria set by the user.
#'
#' The dataset comes as a `list` object, with two elements:
#'
#' * `metadata` - a data frame with metadata regarding drug resources used
#' * `records` - a data frame with drug records
#'
#' @param cache_dir local cache directory for data retrieval
#' @param force_download force download data from remote repository even if
#' data exists in cache
#' @param exclude_salt_forms exclude salt forms of drugs
#' @param exclude_adc exclude antibody-drug conjugates (ADCs)
#' @param treatment_category main treatment category, classified according to
#' ATC or not ('targeted_therapy_classified',
#' 'targeted_therapy_unclassified','chemo_therapy_classified','hormone_therapy_classified',
#' 'immuno_suppressants_classified','other')
#' @param drug_is_approved logical indicating if resulting drug records
#' should contain approved drugs only
#' @param drug_target character vector with drug targets (gene symbols)
#' for drug records included in results
#' @param drug_action_type character vector with drug action types to
#' include in drug record list -
#' possible values "INHIBITOR","AGONIST","MODULATOR","ANTAGONIST",
#' "BLOCKER","ACTIVATOR","BINDING AGENT","OPENER",
#' "STABILISER","CROSS-LINKING AGENT",DISRUPTING AGENT","OTHER"
#' @param drug_action_inhibition logical indicating to only return drug records
#' with inhibitory mechanism-of-action
#' @param drug_indication_main character vector with main tumor types for
#' which drug(s) are indicated. Possible values: "Adrenal Gland","Biliary Tract",
#' "Bladder/Urinary Tract","Bone","Breast","Cervix","CNS/Brain",
#' "Colon/Rectum","Esophagus/Stomach","Eye","Head and Neck","Kidney",
#' "Liver","Lung","Lymphoid","Myeloid","Ovary/Fallopian Tube",
#' "Pancreas","Penis","Peripheral Nervous System","Peritoneum",
#' "Pleura","Prostate","Skin","Soft Tissue","Testis","Thymus",
#' "Thyroid","Uterus","Vulva/Vagina"
#' @param drug_source_opentargets logical indicating if resulting drug records
#' should contain drug records from Open Targets Platform/ChEMBL only
#' @param drug_cancer_indication logical indicating if resulting drug records
#' should be for those indicated for cancer conditions only (from approved
#' conditions, found in clinical trials etc.)
#' @param drug_classified_cancer logical indicating if resulting drug
#' records should be for those classified only in the "L" class of ATC (
#' "ANTINEOPLASTIC AND IMMUNOMODULATING AGENTS") only
#' @param drug_has_blackbox_warning logical indicating if resulting drug
#' records should contain drugs with black box warnings only
#' @param drug_approval_year only include records for drugs approved
#' later than this date (year)
#' @param drug_minimum_phase_any_indication only include drug records that
#' are in a clinical phase (any indication) greater or equal than this phase
#' @param output_resolution dictate output record resolution
#' ('drug','drug2target','drug2target2indication')
#'
#' @return
#'
#' The `records` data frame contains the following columns (only selected columns
#' will be shown based on the value of `output_resolution`)
#'
#' \itemize{
#' \item \emph{drug_id} - drug identifier (pharmaOncoX)
#' \item \emph{drug_name} - primary drug name (upper case, NCI Thesaurus)
#' \item \emph{drug_type} - type of drug molecule (Antibody, small molecule etc)
#' \item \emph{molecule_chembl_id} - ChEMBL compound identifier
#' \item \emph{drug_action_type} - main action elicited by drug
#' (antagonist, inhibitor, stabiliser etc)
#' \item \emph{drug_alias} - collection of unambiguous drug aliases
#' (separated by '|')
#' \item \emph{nci_concept_definition} - detailed description of drug
#' mechanism-of-action (NCI Thesaurus)
#' \item \emph{opentargets} - logical - drug is found in the Open Targets
#' Platform resource
#' \item \emph{is_salt} - logical - drug record represents a salt form
#' (excluded by default)
#' \item \emph{is_adc} - logical - drug record represents an
#' antibody-drug conjugate (ADC - excluded by default)
#' \item \emph{drug_blacbox_warning} - logical indicating if drug has
#' blackbox warning
#' \item \emph{nci_t} - NCI thesaurus identifier
#' \item \emph{target_symbol} - gene symbol of drug target
#' \item \emph{target_entrezgene} - Entrez gene identifier of drug target
#' \item \emph{target_genename} - gene name/description of drug target
#' \item \emph{target_ensembl_gene_id} - Ensembl gene identifier of drug target
#' \item \emph{target_type} - type of drug target (single protein, protein
#' family etc.)
#' \item \emph{drug_max_phase_indication} - maximum clinical phase for drug
#' (given indication)
#' \item \emph{drug_approved_indication} - logical indicating if drug has
#' an approved indication
#' \item \emph{drug_frac_cancer_indications} - fraction of drug indications that
#' are for cancers
#' \item \emph{drug_approved_noncancer} - logical indicating if drug
#' is approved for a non-cancer disease
#' \item \emph{drug_n_indications} - number of indications for the
#' given drug (from approved indications, clinical trials etc)
#' \item \emph{drug_year_first_approval} - year drug was first approved
#' \item \emph{drug_max_ct_phase} - maximum clinical phase for drug
#' (any indication)
#' \item \emph{disease_efo_id} - EFO (Experimental Factor Ontology)
#' identifier for drug indication
#' \item \emph{disease_efo_label} - EFO (Experimental Factor Ontology)
#' label for drug indication
#' \item \emph{primary_site} - primary tumor site/type (obtained through
#' https://github.com/sigven/oncoPhenoMap)
#' \item \emph{drug_clinical_id} - drug clinical identifier
#' (clinicaltrials.gov, DailyMed, FDA etc.)
#' \item \emph{drug_clinical_source} - underlying source for drug entry
#' (DailyMed, clinicaltrials.gov, FDA etc.)
#' \item \emph{atc_code_level1} - drug identifier ATC (level 1)
#' \item \emph{atc_level1} - drug label ATC (level 1)
#' \item \emph{atc_code_level1} - drug identifier ATC (level 2)
#' \item \emph{atc_level2} - drug label ATC (level 2)
#' \item \emph{atc_code_level3} - drug identifier ATC (level 3)
#' \item \emph{atc_level3} - drug label ATC (level 3)
#' \item \emph{atc_treatment_category} - treatment category (targeted/chemo/hormone, cancer/other etc)
#' }
#'
#'
#'
#' @export
#'
get_drugs <- function(
cache_dir = NA,
force_download = FALSE,
exclude_salt_forms = TRUE,
exclude_adc = FALSE,
treatment_category =
c("targeted_therapy_classified",
"targeted_therapy_unclassified",
"chemo_therapy_classified",
"hormone_therapy_classified",
"immuno_suppressants_classified",
"other"),
drug_is_approved = FALSE,
drug_target = NULL,
drug_action_type = NULL,
drug_indication_main = NULL,
drug_source_opentargets = FALSE,
drug_cancer_indication = TRUE,
drug_classified_cancer = TRUE,
drug_has_blackbox_warning = FALSE,
drug_approval_year = 1939,
drug_minimum_phase_any_indication = 0,
output_resolution = "drug2target2indication",
drug_action_inhibition = F){
## set logging layout
lgr::lgr$appenders$console$set_layout(
lgr::LayoutFormat$new(timestamp_fmt = "%Y-%m-%d %T"))
valid_output_resolutions <-
c("drug",
"drug2target",
"drug2target2indication")
valid_treatment_categories <-
c("hormone_therapy_classified",
"targeted_therapy_classified",
"targeted_therapy_unclassified",
"chemo_therapy_classified",
"immuno_suppressants_classified",
"other")
is_valid_treatment_category <-
unique(treatment_category %in% valid_treatment_categories)
valid_drug_action_types <-
c("INHIBITOR",
"AGONIST",
"MODULATOR",
"ANTAGONIST",
"BLOCKER",
"ACTIVATOR",
"BINDING AGENT",
"OPENER",
"STABILISER",
"CROSS-LINKING AGENT",
"DISRUPTING AGENT",
"OTHER")
valid_indications_main <-
c("Adrenal Gland",
"Biliary Tract",
"Bladder/Urinary Tract",
"Bone",
"Breast",
"Cervix",
"CNS/Brain",
"Colon/Rectum",
"Esophagus/Stomach",
"Eye",
"Head and Neck",
"Kidney",
"Liver",
"Lung",
"Lymphoid",
"Myeloid",
"Ovary/Fallopian Tube",
"Pancreas",
"Penis",
"Peripheral Nervous System",
"Peritoneum",
"Pleura",
"Prostate",
"Skin",
"Soft Tissue",
"Testis",
"Thymus",
"Thyroid",
"Uterus",
"Vulva/Vagina")
arg_validation_messages <- list()
arg_validation_messages[[1]] <-
assertthat::validate_that(
is.numeric(drug_approval_year),
msg = "ERROR: Argument 'drug_approval_year' must be of type 'numeric'")
arg_validation_messages[[2]] <-
assertthat::validate_that(
drug_approval_year >= 1939 & drug_approval_year <= 2024,
msg = "ERROR: Argument 'drug_approval_year' must be larger than 1939 and less than or equal to 2024")
arg_validation_messages[[3]] <-
assertthat::validate_that(
is.logical(drug_is_approved),
msg = "ERROR: Argument 'drug_is_approved' must be of type 'logical'")
arg_validation_messages[[4]] <-
assertthat::validate_that(
is.logical(drug_source_opentargets),
msg = "ERROR: Argument 'drug_source_opentargets' must be of type 'logical'")
arg_validation_messages[[5]] <-
assertthat::validate_that(
is.logical(drug_classified_cancer),
msg = "ERROR: Argument 'drug_classified_cancer' must be of type 'logical'")
arg_validation_messages[[6]] <-
assertthat::validate_that(
is.logical(drug_cancer_indication),
msg = "ERROR: Argument 'drug_cancer_indication' must be of type 'logical'")
arg_validation_messages[[7]] <-
assertthat::validate_that(
is.character(treatment_category),
msg = "ERROR: Argument 'treatment_category' must be of type 'character'")
arg_validation_messages[[8]] <-
assertthat::validate_that(
length(is_valid_treatment_category) == 1 &
is_valid_treatment_category == T,
msg = paste0("ERROR: Argument 'treatment_category' must contain any ",
"combination of 'chemo_therapy','hormone_therapy', or 'targeted_therapy'"))
arg_validation_messages[[9]] <-
assertthat::validate_that(
is.logical(drug_has_blackbox_warning),
msg = "ERROR: Argument 'drug_has_blackbox_warning' must be of type 'logical'")
arg_validation_messages[[10]] <-
assertthat::validate_that(
output_resolution %in% valid_output_resolutions,
msg = "ERROR: Argument 'output_resolution' must be either 'drug','drug2target', or 'drug2target2indication'")
arg_counter <- 11
if (!is.null(drug_action_type)) {
arg_counter <- arg_counter + 1
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
is.character(drug_action_type),
msg = "ERROR: Argument 'drug_action_type' must be a vector of type 'character'")
arg_counter <- arg_counter + 1
if (length(unique(drug_action_type %in% valid_drug_action_types)) > 1) {
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
F,
msg = paste0(
"ERROR: Argument 'drug_action_type' must be a character vector",
" with any of the following action types: ",
paste(valid_drug_action_types, collapse = ", ")))
arg_counter <- arg_counter + 1
}
}
if (!is.null(drug_indication_main)) {
#arg_counter <- arg_counter + 1
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
is.character(drug_indication_main),
msg = "ERROR: Argument 'drug_indication_main' must be a vector of type 'character'")
arg_counter <- arg_counter + 1
if (length(unique(drug_indication_main %in% valid_indications_main)) > 1) {
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
F,
msg = paste0(
"ERROR: Argument 'drug_indication_main' must be a character vector",
" with any of the following action types: ",
paste(valid_indications_main, collapse = ", ")))
arg_counter <- arg_counter + 1
}
}
if (!is.null(drug_target)) {
arg_validation_messages[[arg_counter]] <-
assertthat::validate_that(
is.character(drug_target),
msg = "ERROR: Argument 'drug_target' must be a vector of type 'character'")
arg_counter <- arg_counter + 1
}
i <- 1
error_messages <- c()
while (i <= length(arg_validation_messages)) {
if (!is.logical(arg_validation_messages[[i]])) {
error_messages <- c(error_messages, arg_validation_messages[[i]])
}
i <- i + 1
}
if (length(error_messages) > 0) {
lgr::lgr$info('\n')
lgr::lgr$info(error_messages, sep = "\n")
return()
}
## Get full list of drug records
all_drug_recs <- get_drug_records(cache_dir, force_download)
drug_records <- as.data.frame(all_drug_recs[['records']])
metadata <- as.data.frame(all_drug_recs[['metadata']])
if (exclude_salt_forms == TRUE) {
drug_records <- drug_records |>
dplyr::filter(.data$is_salt == FALSE)
}
if (exclude_adc == TRUE) {
drug_records <- drug_records |>
dplyr::filter(.data$is_adc == FALSE)
}
## targeted + chemo + hormone
treatment_category_regex <- ""
atc_treatment_strings <- c()
if("hormone_therapy_classified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer_hormone_therapy")
}
if("chemo_therapy_classified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer_chemo_therapy")
}
if("targeted_therapy_classified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer(_adc)?_targeted_therapy")
}
if("targeted_therapy_unclassified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer_unclassified")
}
if("immuno_suppressants_classified" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "cancer_immuno_suppressants")
}
if("other" %in% treatment_category){
atc_treatment_strings <- c(atc_treatment_strings, "(other_targeted_therapy|unknown)")
}
treatment_category_regex <- paste0(
"^(", paste(atc_treatment_strings, collapse="|"), ")$")
#cat(paste0('Treatment category regex: ', treatment_category_regex))
#cat('\n')
drug_records <- drug_records |>
dplyr::filter(
!is.na(.data$atc_treatment_category) &
stringr::str_detect(
.data$atc_treatment_category,
treatment_category_regex))
if (drug_source_opentargets == TRUE) {
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(.data$opentargets == TRUE)
}
}
if (drug_cancer_indication == TRUE){
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(stringr::str_detect(
.data$drug_cancer_relevance, "by_cancer_"
))
}
}
if (drug_classified_cancer == TRUE){
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(
!is.na(.data$atc_treatment_category) &
stringr::str_detect(
.data$atc_treatment_category, "cancer_"
)
)
}
}
if (drug_action_inhibition == TRUE) {
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(.data$inhibition_moa == TRUE)
}
}
if (drug_has_blackbox_warning == TRUE) {
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(is.na(.data$drug_blackbox_warning) |
.data$drug_blackbox_warning == TRUE)
if (nrow(drug_records) == 0) {
lgr::lgr$info(
paste0("WARNING: For the conditions listed below, NO drugs were found with a blackbox warning\n"))
lgr::lgr$info(
paste0("Condition 1: Drug treatment category: ",
paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(
paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info('\n')
}
}
}
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(is.na(.data$drug_year_first_approval) |
.data$drug_year_first_approval >= drug_approval_year)
if (nrow(drug_records) == 0) {
lgr::lgr$info(
paste0("WARNING: For the conditions listed below, NO drugs were found with an approval date greater than or equal to: ",
drug_approval_year),"\n")
lgr::lgr$info(
paste0("Condition 1: Drug treatment category: ",
paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(
paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info(
paste0("Condition 3: Drugs with blackbox warnings only: ", drug_has_blackbox_warning, "\n"))
lgr::lgr$info('\n')
}
}
if (drug_minimum_phase_any_indication > 0) {
if (nrow(drug_records) > 0) {
## include indications customly retrieved in DailyMed (assuming they are all max phase)
drug_records <- drug_records |>
dplyr::filter(
(!is.na(.data$drug_max_ct_phase) &
.data$drug_max_ct_phase >= drug_minimum_phase_any_indication))
if (nrow(drug_records) == 0) {
lgr::lgr$info(paste0(
"WARNING: For the conditions listed below, NO drugs were found",
"with a clinical phase greater or equal than: ",
drug_minimum_phase_any_indication),"\n")
lgr::lgr$info(
paste0("Condition 1: Drug treatment category: ",
paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(
paste0("Condition 2: Open Targets Platform only: ",
drug_source_opentargets, "\n"))
lgr::lgr$info(
paste0("Condition 3: Drugs with blackbox warnings only: ",
drug_has_blackbox_warning, "\n"))
lgr::lgr$info(
paste0("Condition 4: Drugs with approval later than only: ",
drug_approval_year, "\n"))
lgr::lgr$info('\n')
}
}
}
if (drug_is_approved == TRUE) {
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(!is.na(.data$drug_approved_indication) &
.data$drug_approved_indication == TRUE)
}
if (nrow(drug_records) == 0) {
lgr::lgr$info(
paste0("WARNING: For the conditions listed below, NO approved drugs were found\n"))
lgr::lgr$info(
paste0("Condition 1: Drug treatment category: ",
paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(
paste0("Condition 2: Open Targets Platform only: ",
drug_source_opentargets, "\n"))
lgr::lgr$info(
paste0("Condition 3: Drugs with blackbox warnings only: ",
drug_has_blackbox_warning, "\n"))
lgr::lgr$info(
paste0("Condition 4: Drugs with approval later than only: ",
drug_approval_year, "\n"))
lgr::lgr$info(
paste0("Condition 5: Drugs with minimum clinical phase (any indication) only: ",
drug_minimum_phase_any_indication, "\n"))
}
}
all_drug_targets <- data.frame()
if (!is.null(drug_target)) {
for (target in drug_target) {
df <- data.frame('target_symbol' = target,
stringsAsFactors = F)
all_drug_targets <- dplyr::bind_rows(
all_drug_targets, df
)
}
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::filter(!is.na(.data$target_symbol)) |>
dplyr::inner_join(
all_drug_targets, by = "target_symbol")
if (nrow(drug_records) == 0) {
lgr::lgr$info(paste0("WARNING: For the conditions listed below, NO molecularly targeted drugs were found for the target proteins provided in the 'drug_target' argument: ",
paste(all_drug_targets$target_symbol, collapse = ", "),"\n"))
lgr::lgr$info(paste0("Condition 1: Drug treatment category: ", paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info(paste0("Condition 3: Drugs with approved indications only: ", drug_is_approved, "\n"))
lgr::lgr$info(paste0("Condition 4: Drugs with blackbox warnings only: ", drug_has_blackbox_warning, "\n"))
lgr::lgr$info(paste0("Condition 5: Drugs with approval later than only: ", drug_approval_year, "\n"))
lgr::lgr$info(paste0("Condition 6: Drugs with minimum clinical phase (any indication) only: ", drug_minimum_phase_any_indication, "\n"))
lgr::lgr$info('\n')
}else{
drug_records <- drug_records |>
dplyr::arrange(
dplyr::desc(.data$drug_max_ct_phase),
dplyr::desc(.data$drug_year_first_approval))
}
}
}
if (!is.null(drug_action_type)) {
if (nrow(drug_records) > 0) {
drug_records_action_type <- data.frame()
drug_records$atype <- drug_records$drug_action_type
for (actype in drug_action_type) {
recs <- drug_records |>
dplyr::filter(
!is.na(.data$atype) &
stringr::str_detect(.data$atype, actype)
)
if (nrow(recs) > 0) {
drug_records_action_type <- drug_records_action_type |>
dplyr::bind_rows(recs)
}
}
if (nrow(drug_records_action_type) == 0) {
lgr::lgr$info(paste0("WARNING: For the conditions listed below, no drug records were found for the action types provided in the argument 'drug_action_type': "
, paste(drug_action_type, collapse = ", ")),"\n")
lgr::lgr$info(paste0("Condition 1: Drug treatment category: ", paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info(paste0("Condition 3: Drugs with approved indications only: ", drug_is_approved, "\n"))
lgr::lgr$info(paste0("Condition 4: Drugs with blackbox warnings only: ", drug_has_blackbox_warning, "\n"))
lgr::lgr$info(paste0("Condition 5: Drugs with approval later than: ", drug_approval_year, "\n"))
lgr::lgr$info(paste0("Condition 6: Drugs with minimum clinical phase (any indication) only: ", drug_minimum_phase_any_indication, "\n"))
if (nrow(all_drug_targets) > 0) {
lgr::lgr$info(paste0("Condition 7: Drugs targeted towards the following targets: ", paste(all_drug_targets$target_symbol, collapse = ", "), "\n"))
}
lgr::lgr$info('\n')
}else{
drug_records <- drug_records_action_type |>
dplyr::select(-.data$atype)
}
}
}
if (!is.null(drug_indication_main)) {
if (nrow(drug_records) > 0) {
drug_records_indication <- data.frame()
for (ind in drug_indication_main) {
recs <- drug_records |>
dplyr::filter(
!is.na(.data$primary_site) &
stringr::str_detect(.data$primary_site, ind)
)
if (nrow(recs) > 0) {
drug_records_indication <- drug_records_indication |>
dplyr::bind_rows(recs)
}
}
if (nrow(drug_records_indication) == 0) {
lgr::lgr$info(paste0("WARNING: For the conditions listed below, no drug records were found for the indications (main tumor types) provided in the argument 'drug_indication_main': "
, paste(drug_indication_main, collapse = ", ")),"\n")
lgr::lgr$info(paste0("Condition 1: Drug treatment category: ", paste(treatment_category, collapse=", "), "\n"))
lgr::lgr$info(paste0("Condition 2: Open Targets Platform only: ", drug_source_opentargets, "\n"))
lgr::lgr$info(paste0("Condition 3: Drugs with approved indications only: ", drug_is_approved, "\n"))
lgr::lgr$info(paste0("Condition 4: Drugs with blackbox warnings only: ", drug_has_blackbox_warning, "\n"))
lgr::lgr$info(paste0("Condition 5: Drugs with approval later than: ", drug_approval_year, "\n"))
lgr::lgr$info(paste0("Condition 6: Drugs with minimum clinical phase (any indication) only: ", drug_minimum_phase_any_indication, "\n"))
if (nrow(all_drug_targets) > 0) {
lgr::lgr$info(paste0("Condition 7: Drugs targeted towards the following targets only: ", paste(all_drug_targets$target_symbol, collapse = ", "), "\n"))
}
if (!is.null(drug_action_type)) {
lgr::lgr$info(paste0(
"Condition 8: Drugs with the following action types only: ",
paste(drug_action_type, collapse = ", "), "\n"))
}
lgr::lgr$info('\n')
}else{
drug_records <- drug_records_indication
}
}
}
lgr::lgr$info(
paste0("Record set satisfying user-defined criteria: n = ",
nrow(drug_records)))
if (output_resolution == "drug") {
lgr::lgr$info(
"Collapsing record set - providing output on a 'per_drug' resolution")
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::select(-c(.data$target_symbol,
.data$target_ensembl_gene_id,
.data$target_entrezgene,
.data$target_type,
.data$target_genename,
.data$disease_efo_id,
.data$cui,
.data$cui_name,
.data$drug_approved_indication,
.data$drug_clinical_source)) |>
dplyr::distinct()
drug_records <- as.data.frame(
drug_records |>
dplyr::group_by(
dplyr::across(-c(
.data$atc_code_level1,
.data$atc_level1,
.data$atc_code_level2,
.data$atc_level2,
.data$atc_code_level3,
.data$atc_level3,
.data$atc_treatment_category,
.data$disease_efo_label,
.data$primary_site,
.data$drug_clinical_id,
.data$drug_max_phase_indication))) |>
dplyr::summarise(
drug_clinical_id = paste(
unique(sort(.data$drug_clinical_id)), collapse = "|"),
disease_indication = paste(
unique(sort(.data$disease_efo_label)), collapse = "|"),
disease_indication_max_phase = paste(
sort(unique(.data$drug_max_phase_indication),
decreasing = T), collapse = "|"),
disease_main_group = paste(
unique(sort(.data$primary_site)), collapse = "|"),
atc_code_level1 = paste(
unique(sort(.data$atc_code_level1)),
collapse = "|"
),
atc_level1 = paste(
unique(sort(.data$atc_level1)),
collapse = "|"
),
atc_code_level2 = paste(
unique(sort(.data$atc_code_level2)),
collapse = "|"
),
atc_level2 = paste(
unique(sort(.data$atc_level2)),
collapse = "|"
),
atc_code_level3 = paste(
unique(sort(.data$atc_code_level3)),
collapse = "|"
),
atc_level3 = paste(
unique(sort(.data$atc_level3)),
collapse = "|"
),
atc_treatment_category = paste(
unique(sort(.data$atc_treatment_category)),
collapse = "|"
),
.groups = "drop") |>
dplyr::distinct()
)
}
lgr::lgr$info(
paste0("Final record set: n = ", nrow(drug_records), " records"))
}
if (output_resolution == "drug2target") {
lgr::lgr$info("Collapsing record set - providing output on a 'per_drug_target' resolution")
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::select(-c(.data$disease_efo_id,
.data$cui,
.data$cui_name,
.data$drug_approved_indication,
.data$drug_clinical_source)) |>
dplyr::distinct()
drug_records <- as.data.frame(
drug_records |>
dplyr::group_by(
dplyr::across(
-c(.data$disease_efo_label,
.data$atc_code_level1,
.data$atc_level1,
.data$atc_code_level2,
.data$atc_level2,
.data$atc_code_level3,
.data$atc_level3,
.data$atc_treatment_category,
.data$primary_site,
.data$drug_clinical_id,
.data$drug_max_phase_indication))) |>
dplyr::summarise(
drug_clinical_id = paste(
unique(sort(.data$drug_clinical_id)), collapse = "|"),
disease_indication = paste(
unique(sort(.data$disease_efo_label)), collapse = "|"),
disease_indication_max_phase = paste(
sort(unique(.data$drug_max_phase_indication),
decreasing = F), collapse = "|"),
disease_main_group = paste(
unique(sort(.data$primary_site)), collapse = "|"),
atc_code_level1 = paste(
unique(sort(.data$atc_code_level1)),
collapse = "|"
),
atc_level1 = paste(
unique(sort(.data$atc_level1)),
collapse = "|"
),
atc_code_level2 = paste(
unique(sort(.data$atc_code_level2)),
collapse = "|"
),
atc_level2 = paste(
unique(sort(.data$atc_level2)),
collapse = "|"
),
atc_code_level3 = paste(
unique(sort(.data$atc_code_level3)),
collapse = "|"
),
atc_level3 = paste(
unique(sort(.data$atc_level3)),
collapse = "|"
),
atc_treatment_category = paste(
unique(sort(.data$atc_treatment_category)),
collapse = "|"
),
.groups = "drop") |>
dplyr::distinct()
)
}
lgr::lgr$info(
paste0("Final record set: n = ", nrow(drug_records), " records"))
}
if (output_resolution == "drug2target2indication") {
lgr::lgr$info(
"Collapsing record set - providing output on a 'per_drug_target_indication' resolution")
if (nrow(drug_records) > 0) {
drug_records <- as.data.frame(
drug_records |>
dplyr::select(-c(.data$cui, .data$cui_name)) |>
dplyr::distinct() |>
dplyr::group_by(
dplyr::across(
-c(.data$drug_clinical_id,
.data$atc_code_level1,
.data$atc_level1,
.data$atc_code_level2,
.data$atc_level2,
.data$atc_code_level3,
.data$atc_level3,
.data$atc_treatment_category,
.data$drug_clinical_source))) |>
dplyr::summarise(
drug_clinical_id = paste(
unique(sort(.data$drug_clinical_id)), collapse = "|"),
drug_clinical_source = paste(
unique(sort(.data$drug_clinical_source)), collapse = "|"),
atc_code_level1 = paste(
unique(sort(.data$atc_code_level1)),
collapse = "|"
),
atc_level1 = paste(
unique(sort(.data$atc_level1)),
collapse = "|"
),
atc_code_level2 = paste(
unique(sort(.data$atc_code_level2)),
collapse = "|"
),
atc_level2 = paste(
unique(sort(.data$atc_level2)),
collapse = "|"
),
atc_code_level3 = paste(
unique(sort(.data$atc_code_level3)),
collapse = "|"
),
atc_level3 = paste(
unique(sort(.data$atc_level3)),
collapse = "|"
),
atc_treatment_category = paste(
unique(sort(.data$atc_treatment_category)),
collapse = "|"
),
.groups = "drop") |>
dplyr::distinct()
)
}
lgr::lgr$info(
paste0("Final record set: n = ", nrow(drug_records), " records"))
}
if (nrow(drug_records) > 0) {
drug_records <- drug_records |>
dplyr::select(
.data$drug_id,
.data$drug_name,
.data$drug_type,
.data$molecule_chembl_id,
.data$drug_action_type,
.data$drug_alias,
.data$nci_concept_definition,
.data$opentargets,
dplyr::everything()
) |>
dplyr::arrange(
dplyr::desc(.data$drug_max_ct_phase),
dplyr::desc(.data$drug_year_first_approval),
dplyr::desc(.data$opentargets),
nchar(.data$drug_name))
if(!is.null(drug_indication_main)){
if(output_resolution != "drug2target2indication"){
drug_records <- drug_records |>
dplyr::arrange(
dplyr::desc(.data$disease_indication_max_phase),
dplyr::desc(.data$drug_year_first_approval),
dplyr::desc(.data$opentargets),
nchar(.data$drug_name))
}
}
}
oncodrugs <- list()
oncodrugs[['records']] <- drug_records
oncodrugs[['metadata']] <- metadata
return(oncodrugs)
}
#' Get curated cancer biomarker datasets
#'
#' @description
#' Downloads preprocessed datasets to a local cache directory and returns a
#' curated set of genomic biomarkers from multiple sources
#' (CIViC, CGI, MitelmanDB)
#'
#' The dataset comes as a `list` object, with three elements:
#'
#' * `metadata` - a data frame with metadata regarding drug resources used
#' * `data` - a list with four elements ('civic','cgi','mitelmandb','custom_fusions')
#' * `fpath` - path to cache file
#'
#' @param cache_dir Local directory for data download
#' @param force_download Logical indicating if local cache should force downloaded
#' (i.e. set to TRUE to re-download even if data exists in cache)
#'
#' @return
#'
#' Each entry of the source-specific (e.g. 'civic') entry in the `data` list contains
#' a list of three data frames:
#'
#' \itemize{
#' \item \emph{variant} - list of all biomarker variants, extensively populated
#' according to variant aliases (identifer - column \strong{variant_id})
#' \item \emph{clinical} - cross-references between variants recorded in
#' the `variant` data frame and clinical evidence items (identifier -
#' column \strong{evidence_id}) and underlying literature evidence
#' (identifier - column \strong{source_id})
#' \item \emph{literature} - lists literature for all source_id's listed in
#' the `clinical` data frame
#' }
#'
#' @export
#'
get_biomarkers <- function(cache_dir = NA,
force_download = F) {
lgr::lgr$appenders$console$set_layout(
lgr::LayoutFormat$new(timestamp_fmt = "%Y-%m-%d %T"))
if (is.na(cache_dir)) {
lgr::lgr$fatal(paste0("Argument cache_dir = '",
cache_dir, "' is not defined"))
stop()
}
if (!dir.exists(cache_dir)) {
lgr::lgr$fatal(paste0("Argument cache_dir = '",
cache_dir, "' does not exist"))
stop()
}
biomarker_datasets <- list()
file_maps <- c('biomarkers')
for (elem in file_maps) {
fname_local <- file.path(
cache_dir,
paste0(elem,"_v",
db_id_ref[db_id_ref$name == elem,]$pVersion,
'.rds')
)
fname_gd <- googledrive::as_id(
db_id_ref[db_id_ref$name == elem,]$gid)
md5checksum_package <-
db_id_ref[db_id_ref$name == elem,]$md5Checksum
#dat <- NULL
if (file.exists(fname_local) & force_download == F) {
biomarker_datasets[[elem]] <- readRDS(fname_local)
biomarker_datasets[[elem]][['fpath']] <- fname_local
if (!is.null(biomarker_datasets[[elem]][['data']]) &
!is.null(biomarker_datasets[[elem]][['metadata']])) {
lgr::lgr$info(paste0(
"Reading from cache_dir = '",
cache_dir, "', argument force_download = F"))
lgr::lgr$info(paste0("Object '",elem,"' sucessfully loaded"))
}
}else{
googledrive::drive_deauth()
lgr::lgr$info("Downloading remote dataset from Google Drive to cache_dir")
dl <- googledrive::with_drive_quiet(
googledrive::drive_download(
fname_gd,
path = fname_local,
overwrite = TRUE)
)
md5checksum_remote <- dl$drive_resource[[1]]$md5Checksum
md5checksum_local <- tools::md5sum(fname_local)
names(md5checksum_local) <- NULL
if (md5checksum_remote == md5checksum_local) {
biomarker_datasets[[elem]] <- readRDS(fname_local)
biomarker_datasets[[elem]]$fpath <- fname_local
if (!is.null(biomarker_datasets[[elem]][['data']]) &
!is.null(biomarker_datasets[[elem]][['metadata']])) {
lgr::lgr$info(paste0(
"Reading from cache_dir = ' (",
cache_dir, "'), argument force_download = F"))
lgr::lgr$info(paste0("Object '", elem, "' sucessfully loaded"))
lgr::lgr$info(paste0("md5 checksum is valid: ", md5checksum_remote))
}
}else{
lgr::lgr$error(paste0("md5 checksum of local file (", md5checksum_local,
") is inconsistent with remote file (",
md5checksum_remote,")"))
stop()
}
}
}
return(biomarker_datasets[['biomarkers']])
}
#' Tidy eval helpers
#'
#' <https://cran.r-project.org/web/packages/dplyr/vignettes/programming.html>
#'
#' @name tidyeval
#' @keywords internal
#' @importFrom rlang .data :=
NULL
utils::globalVariables(c("."))
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.