Merge pull request #156 from umccr/piedb-comp

PieDB comparison

README.md

@@ -371,7 +371,7 @@ dataset is used as reference cohort (`--dataset TEST`).
 rnasum.R \
   --sample_name test_sample_WTS \
   --dataset TEST \
-  --dragen_rnaseq inst/rawdata/test_data/dragen \
+  --dragen_wts_dir inst/rawdata/test_data/dragen \
   --report_dir inst/rawdata/test_data/dragen/RNAsum \
   --umccrise inst/rawdata/test_data/umccrised/test_sample_WGS \
   --save_tables FALSE
@@ -396,7 +396,7 @@ pancreatic adenocarcinoma dataset is used as the reference cohort
 rnasum.R \
   --sample_name test_sample_WTS \
   --dataset TEST \
-  --dragen_rnaseq inst/rawdata/test_data/dragen \
+  --dragen_wts_dir inst/rawdata/test_data/dragen \
   --report_dir inst/rawdata/test_data/dragen/RNAsum \
   --save_tables FALSE
 ```
@@ -422,7 +422,7 @@ dataset is used as the reference cohort (`--dataset TEST`).
 rnasum.R \
   --sample_name test_sample_WTS \
   --dataset TEST \
-  --dragen_rnaseq $(pwd)/../rawdata/test_data/dragen \
+  --dragen_wts_dir $(pwd)/../rawdata/test_data/dragen \
   --report_dir $(pwd)/../rawdata/test_data/dragen/RNAsum \
   --umccrise $(pwd)/../rawdata/test_data/umccrised/test_sample_WGS \
   --save_tables FALSE \

inst/scripts/compare_piedb.R

@@ -0,0 +1,50 @@
+require(dplyr)
+require(readr)
+require(here)
+require(glue)
+require(tibble)
+require(tidyr)
+require(DT)
+
+SBJ04426_dev <- readr::read_csv("../../data/wts/RNAsum/reference_update_comp/SBJ04426/dev/genes.expr.perc.csv")
+SBJ04426_pro <- readr::read_csv("../../data/wts/RNAsum/reference_update_comp/SBJ04426/pro/genes.expr.perc.csv")
+SBJ04187_dev <- readr::read_csv("../../data/wts/RNAsum/reference_update_comp/SBJ04187/dev/genes.expr.perc.csv")
+SBJ04187_pro <- readr::read_csv("../../data/wts/RNAsum/reference_update_comp/SBJ04187/pro/genes.expr.perc.csv")
+cancer_genes <- readr::read_tsv("../../research/data/cancer_gene_list/somatic_panel-v24.03.0.tsv")
+
+# now explore expression differences in reference and patient columns
+# between dev and prod.
+SBJ04426_df <- dplyr::left_join(SBJ04426_dev, SBJ04426_pro, by = "Gene", suffix = c(".dev", ".pro")) |>
+  dplyr::mutate(
+    Ref_equal = `BRCA (TCGA).dev` == `BRCA (TCGA).pro`,
+    Pat_equal = Patient.dev == Patient.pro,
+    # Ref_diff = abs(`PANCAN (TCGA).dev` - `PANCAN (TCGA).pro`),
+    Ref_diff = abs(`BRCA (TCGA).dev` - `BRCA (TCGA).pro`),
+    Pat_diff = abs(Patient.dev - Patient.pro)
+  ) |>
+  dplyr::select(Gene, contains("BRCA"), contains("PANCAN"), Ref_diff, contains("Patient"), Pat_diff, contains("equal")) |>
+  dplyr::filter(Pat_diff > 0) |>
+  dplyr::filter(Gene %in% cancer_genes$ensembl_gene_symbol) |>
+  dplyr::arrange(desc(Pat_diff)) |>
+  dplyr::arrange(desc(Ref_diff)) |>
+  datatable()
+
+SBJ04187_df <- dplyr::left_join(SBJ04187_dev, SBJ04187_pro, by = "Gene", suffix = c(".dev", ".pro")) |>
+  dplyr::mutate(
+    Ref_equal = `BRCA (TCGA).dev` == `BRCA (TCGA).pro`,
+    Pat_equal = Patient.dev == Patient.pro,
+    # Ref_diff = abs(`PANCAN (TCGA).dev` - `PANCAN (TCGA).pro`),
+    Ref_diff = abs(`BRCA (TCGA).dev` - `BRCA (TCGA).pro`),
+    Pat_diff = abs(Patient.dev - Patient.pro)
+  ) |>
+  dplyr::select(Gene, contains("BRCA"), contains("PANCAN"), Ref_diff, contains("Patient"), Pat_diff, contains("equal")) |>
+  dplyr::filter(Pat_diff > 0) |>
+  dplyr::filter(Gene %in% cancer_genes$ensembl_gene_symbol) |>
+  dplyr::arrange(desc(Pat_diff)) |>
+  dplyr::arrange(desc(Ref_diff)) |>
+  datatable()
+
+# plot Ref_diff values
+hist(SBJ04426_df[[1]]$data$Ref_diff, breaks = 100)
+
+