Update ATRT subtyping to use minimal set of genes

analyses/molecular-subtyping-ATRT/01-ATRT-molecular-subtyping-data-prep.Rmd

@@ -198,51 +198,37 @@ collapsed_metadata %>%
 
 ```{r}
 # Define target overexpressed gene vectors
+# https://github.com/AlexsLemonade/OpenPBTA-analysis/issues/244#issuecomment-576850172
 tyr_genes <-
   c("TYR",
     "MITF",
     "DCT",
     "VEGFA",
     "DNAH11",
     "SPEF1",
-    "POU3F4",
-    "POU3F2",
-    "PBX1")
+    "MSX2",
+    "STAT3",
+    "PRRX1",
+    "LMX1",
+    "OTX2")
 shh_genes <-
   c(
     "MYCN",
     "GLI2",
     "CDK6",
     "ASCL1",
-    "HES5/6",
-    "DLL1/3",
-    "ZBTB7A",
-    "RXF3",
-    "RXF2",
-    "MYBL2",
-    "MXI1",
-    "MEIS3",
-    "MEIS2",
-    "MAX",
-    "INSM1",
-    "FOXK1"
+    "HES5",
+    "HES6",
+    "DLL1",
+    "DLL3",
+    "LHX2",
+    "TEAD1"
   )
 myc_genes <-
   c(
     "MYC",
     "HOTAIR",
-    "HOX",
-    "TCF7L2",
-    "STAT1",
-    "REST",
-    "RARG",
-    "RAD21",
-    "NR4A2",
-    "IRF9",
-    "IRF8",
-    "FOXC1",
-    "CEBPB",
-    "ATF4"
+    "TEAD3"
   )
 
 # Filter to only the genes of interest
@@ -394,6 +380,12 @@ rm(gistic_df, atrt_expression_cn_tmb_df)
 # Save final table of results
 
 ```{r}
+# For reordering the output, we will use the vector of genes as input but we 
+# need to account for genes that are missing from the expression matrix
+tyr_genes <- intersect(colnames(final_df), tyr_genes)
+shh_genes <- intersect(colnames(final_df), shh_genes)
+myc_genes <- intersect(colnames(final_df), myc_genes)
+
 # Save final data.frame
 final_df <- final_df %>%
   dplyr::select(
@@ -407,31 +399,13 @@ final_df <- final_df %>%
     location_summary,
     chr_22q_loss,
     SMARCB1_focal_status,
-    TYR,
-    MITF,
-    DCT,
-    VEGFA,
-    DNAH11,
-    SPEF1,
-    POU3F4,
-    POU3F2,
-    PBX1,
+    !!! rlang::syms(tyr_genes),
     SMARCA4_focal_status,
     HALLMARK_NOTCH_SIGNALING,
-    MYCN,
-    GLI2,
-    CDK6,
-    ASCL1,
-    ZBTB7A,
-    MYBL2,
-    MXI1,
-    MEIS3,
-    MEIS2,
-    MAX,
-    INSM1,
-    FOXK1,
+    !!! rlang::syms(shh_genes),
     HALLMARK_MYC_TARGETS_V1,
     HALLMARK_MYC_TARGETS_V2,
+    !!! rlang::syms(myc_genes),
     dplyr::everything()
   )