Merge pull request #96 from gdrplatform/GDR-2829

feat: add support for diferent run in public PRISM
gdrplatform · Jan 22, 2025 · 8da3f28 · 8da3f28
2 parents eb06334 + b9266b5
commit 8da3f28
Show file tree

Hide file tree

Showing 7 changed files with 27 additions and 8 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,8 +1,8 @@
 Package: gDRimport
 Type: Package
 Title: Package for handling the import of dose-response data
-Version: 1.5.4
-Date: 2024-12-09
+Version: 1.5.5
+Date: 2025-01-11
 Authors@R: c(
     person("Arkadiusz", "Gladki", role=c("aut", "cre"), email="[email protected]",
            comment = c(ORCID = "0000-0002-7059-6378")),

diff --git a/NEWS.md b/NEWS.md
@@ -1,3 +1,6 @@
+## gDRimport 1.5.5 - 2025-01-11
+* take into account run in public PRISM data
+
 ## gDRimport 1.5.4 - 2024-12-09
 * support Duration in old private PRISM data format
 

diff --git a/R/package.R b/R/package.R
@@ -16,11 +16,13 @@ DICTIONARY <- list(
 if (getRversion() >= "2.15.1") {
   utils::globalVariables(
     c(
+      "..with",
+      ".N",
       ".SD",
       "Dose",
       "column_name",
       "name",
-      "..with",
+      "new_name",
       "value"
     ),
     utils::packageName())

diff --git a/R/prism_to_gdrDF.R b/R/prism_to_gdrDF.R
@@ -140,6 +140,14 @@ convert_LEVEL6_prism_to_gDR_input <- function(prism_data_path,
                        c("profile_id", "SampleID", "CompoundName", "GeneSymbolOfTargets"),
                        c("column_name", "column_name", "name", "moa"),
                        skip_absent = TRUE)
+
+  # take into account different runs
+  treatment[, new_name := if (.N > 1) paste0(name, "_run",
+                                             data.table::frank(column_name,
+                                                               ties.method = "dense")) else as.character(name),
+            by = name]
+  treatment[, name := new_name]
+  treatment[, new_name := NULL]
 
   checkmate::assert_names(names(cell_lines), must.include = c("row_name",
                                                               "ccle_name"))
@@ -177,6 +185,8 @@ convert_LEVEL6_prism_to_gDR_input <- function(prism_data_path,
   full_data$value <- pmin(readout_min, 2 ^ full_data$value)
   full_data <- full_data[!(is.na(name) | is.na(value))]
 
+
+
   # data for conc = 0
   untrt_tag <- gDRutils::get_env_identifiers("untreated_tag")[1]
   dt_ctrl <- data.table::data.table(clid = unique(full_data$clid),

diff --git a/inst/testdata/prism_collapsed_LOGFC.csv b/inst/testdata/prism_collapsed_LOGFC.csv
@@ -1,2 +1,2 @@
-V1,COL_ID1
-some_depmap_id,-0.0155766399108
+V1,COL_ID1,COL_ID2
+some_depmap_id,-0.0155766399108,-0.1
diff --git a/inst/testdata/prism_treatment.csv b/inst/testdata/prism_treatment.csv
@@ -1,2 +1,3 @@
-column_name,broad_id,name,dose,screen_id,moa,target,disease.area,indication,smiles,phase
-COL_ID1,some_drug_id,some_drug_name,2.325889319,MTS004,some_drug_moa,,,,,Phase X
+column_name,broad_id,name,dose,screen_id,moa,target,disease.area,indication,smiles,phase,profile_id
+COL_ID1,some_drug_id,some_drug_name,2.325889319,MTS004,some_drug_moa,,,,,Phase X,test1
+COL_ID2,some_drug_id,some_drug_name,2.325889319,MTS004,some_drug_moa,,,,,Phase X,test2
diff --git a/tests/testthat/test-prism_to_gdrDF.R b/tests/testthat/test-prism_to_gdrDF.R
@@ -26,8 +26,11 @@ test_that("prism level6  data can be processed into gDR format ", {
                                                                 cell_line_data_path,
                                                                 treatment_data_path)
   expect_is(df_prism$result, "data.table")
-  expect_equal(dim(df_prism$result), c(2, 8))
+  expect_equal(dim(df_prism$result), c(3, 8))
   expect_true(all(gDRutils::get_env_identifiers(c("drug", "cellline"),
                                                 simplify = FALSE) %in% names(df_prism$result)))
+  expect_equal(df_prism[["result"]][[gDRutils::get_env_identifiers("drug")]],
+               c("some_drug_name_run1", "some_drug_name_run2", "vehicle"))
+
 })