get_biomarkers, NCI update

sigven · Sep 1, 2023 · 9b0cd4a · 9b0cd4a
1 parent c7074da
commit 9b0cd4a
Show file tree

Hide file tree

Showing 14 changed files with 302 additions and 125 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,20 +1,24 @@
 Package: pharmOncoX
 Type: Package
-Title: Annotated targeted and non-targeted anticancer drugs
+Title: Molecularly targeted cancer drugs and biomarkers
 Version: 1.4.3
-Date: 2023-08-07
+Date: 2023-09-01
 Authors@R: 
     c(person(given = "Sigve",
              family = "Nakken",
              role = c("aut", "cre"),
              email = "[email protected]",
              comment = c(ORCID = "0000-0001-8468-2050")))
 Maintainer: Sigve Nakken <[email protected]>
-Description: This data package collects anticancer drug information from multiple resources,
-    including Open Targets Platform/ChEMBL, NCI Thesaurus, and PubChem. The main dataset
-    contains a list of all anticancer drugs per indication/tumor type (where this is provided).
-    which can be queried with a range of parameters, including clinical development phase, approval
-    status, drug targets, mechanism-of-action/drug category etc.
+Description: This data package collects anticancer drug information from 
+    multiple resources, including Open Targets Platform/ChEMBL, NCI Thesaurus, 
+    and PubChem. The main dataset contains a list of all molecularly targeted
+    anticancer drugs per indication/tumor type (where this is provided), 
+    which can be queried with a range of parameters, including 
+    clinical development phase, approval status, drug targets, 
+    mechanism-of-action/drug category etc. The package 
+    also allows for the retrieval of curated biomarkers from multiple 
+    freely available resources (CIViC, CGI, Mitelman database).
 License: MIT + file LICENSE
 URL: https://github.com/sigven/pharmOncoX
 BugReports: https://github.com/sigven/pharmOncoX/issues

diff --git a/NAMESPACE b/NAMESPACE
@@ -1,5 +1,6 @@
 # Generated by roxygen2: do not edit by hand
 
+export(get_biomarkers)
 export(get_drugs)
 importFrom(rlang,":=")
 importFrom(rlang,.data)
diff --git a/NEWS.md b/NEWS.md
@@ -1,3 +1,10 @@
+# Version 1.4.4 (September 1st 2023)
+
+* Updated NCI Thesaurus - 23.08d
+* Updated CIViC data
+* Updated Mitelman database (20230803)
+* `get_biomarkers()` now exported as a main function
+
 # Version 1.4.3 (August 7th 2023)
 
 * NCI Thesaurus 23.07e

diff --git a/R/helpers.R b/R/helpers.R
@@ -413,110 +413,4 @@ get_drug_records <- function(cache_dir = NA,
 }
 
 
-#' Function that retrieves pharmOncoX biomarker data (CIViC, CGI, MitellmanDB) 
-#' from Google Drive
-#'
-#' @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)
-#'
-#' @keywords internal
-#'
-#'
-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()
-      }
-
-    }
-  }
-
-  #biomarker_data <- biomarker_datasets[['biomarkers']]
-
-  return(biomarker_datasets[['biomarkers']])
-
-}
-
 
diff --git a/R/pharm_oncox.R b/R/pharm_oncox.R
@@ -144,7 +144,9 @@ get_drugs <- function(
     lgr::LayoutFormat$new(timestamp_fmt = "%Y-%m-%d %T"))
 
   valid_output_resolutions <- 
-    c("drug","drug2target","drug2target2indication")
+    c("drug",
+      "drug2target",
+      "drug2target2indication")
   valid_drug_action_types <-
     c("INHIBITOR",
       "AGONIST",
@@ -846,6 +848,138 @@ get_drugs <- function(
 }
 
 
+
+#' 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>

diff --git a/R/sysdata.rda b/R/sysdata.rda
diff --git a/README.md b/README.md
@@ -2,7 +2,7 @@
 
 # pharmOncoX <a href="https://sigven.github.io/pharmOncoX/"><img src="man/figures/logo.png" align="right" height="130" width="113"/></a>
 
-**pharmOncoX** is an R package that provides access to targeted and non-targeted cancer drugs, including comprehensive annotations per target, drug mechanism-of-action, approval dates, clinical trial phases for various indications etc. Drugs are further classified according to the [Anatomical Therapeutic Chemical (ATC) Classification System](https://www.whocc.no/atc_ddd_index/), enabling a filtering of cancer drugs according to their main types of action.
+**pharmOncoX** is an R package that provides access to targeted and non-targeted cancer drugs, and genomic cancer biomarkers. Cancer drugs include comprehensive annotations per target, drug mechanism-of-action, approval dates, clinical trial phases for various indications etc. Drugs are further classified according to the [Anatomical Therapeutic Chemical (ATC) Classification System](https://www.whocc.no/atc_ddd_index/), enabling a filtering of cancer drugs according to their main types of action.
 
 
 ## Getting started

diff --git a/data-raw/biomarker_utilities.R b/data-raw/biomarker_utilities.R
@@ -2578,10 +2578,10 @@ load_mitelman_db <- function(cache_dir = NA) {
     dplyr::distinct() |>
     dplyr::rename(citation_id = Pubmed)
 
-  mbca_data <- read.table(
+  mbca_data <- readr::read_tsv(
     file = file.path(
       cache_dir, "mitelmandb", "MBCA.TXT.DATA"),
-    sep = "\t", stringsAsFactors = F, header = T) |>
+    show_col_types = F) |>
     dplyr::filter(stringr::str_detect(GeneShort,"::")) |>
     dplyr::rename(variant = GeneShort) |>
     dplyr::rename(karyotype = KaryShort) |>

diff --git a/data-raw/custom_drug_target_regex_nci.tsv b/data-raw/custom_drug_target_regex_nci.tsv
@@ -7,6 +7,9 @@ KTX-1001	NSD2
 Anti-LGR5 Monoclonal Antibody	LGR5
 Apamistamab	PTPRC
 Garsorasib	KRAS
+Anti-BTLA	BTLA
+Cifurtilimab	CD40
+MRTX1133	KRAS
 Izalontamab	HER1
 Izalontamab	ERBB3
 ALDH1/3 Inhibitor ABD-3001	ALDH1

diff --git a/data-raw/drug_utilities.R b/data-raw/drug_utilities.R
@@ -1754,7 +1754,7 @@ map_curated_targets <- function(gene_info = NULL,
     # ) |>
     dplyr::filter(
       stringr::str_detect(
-        nci_concept_definition, "antineoplastic"
+        nci_concept_definition, "antineoplastic|tumor|cancer"
       )
     ) |>
     dplyr::select(nci_cd_name,

diff --git a/data-raw/metadata_pharm_oncox.xlsx b/data-raw/metadata_pharm_oncox.xlsx