fix(getSubnetworkFromIndra): Fix bugs related to multiple matches for an HGNC ID

R/utils_annotateProteinInfoFromIndra.R

@@ -11,13 +11,27 @@
         stop("Input must be a list.")
     }
 
-    if (any(!sapply(uniprotMnemonicIds, is.character))) {
-        stop("All elements in the list must be character strings representing UniProt mnemonic IDs.")
-    }
-
     if (length(uniprotMnemonicIds) == 0) {
         stop("Input list must not be empty.")
     }
+
+    tryCatch({
+        # Attempt to convert all elements to character if not already character
+        uniprotMnemonicIds <- lapply(uniprotMnemonicIds, function(x) {
+            if (!is.character(x)) {
+                as.character(x)
+            } else {
+                x
+            }
+        })
+
+        # Check if conversion was successful
+        if (any(!sapply(uniprotMnemonicIds, is.character))) {
+            stop("All elements in the list must be character strings representing UniProt mnemonic IDs.")
+        }
+    }, error = function(e) {
+        stop("An error occurred converting uniprot mnemonic IDs to character strings: ", e$message)
+    })
 
     apiUrl <- file.path(Sys.getenv("INDRA_API_URL"), "api/get_uniprot_ids_from_uniprot_mnemonic_ids")
 

R/utils_getSubnetworkFromIndra.R

@@ -7,7 +7,7 @@
         stop("Invalid Input Error: Input must contain a column named 'HgncId'.")
     }
     if (nrow(input) >= 400) {
-        stop("Invalid Input Error: INDRA query must contain less than 400 proteins.  Consider adding a p-value cutoff")
+        stop("Invalid Input Error: INDRA query must contain less than 400 proteins.  Consider lowering your p-value cutoff")
     }
 }
 
@@ -64,8 +64,24 @@
         edge$target_id, "@HGNC&format=html",
         sep = ""
     )
-    edge$source_uniprot_id <- input[input$HgncId == edge$source_id, ]$Protein
-    edge$target_uniprot_id <- input[input$HgncId == edge$target_id, ]$Protein
+
+    # Convert back to uniprot IDs
+    matched_rows_source <- input[input$HgncId == edge$source_id & !is.na(input$Protein), ]
+    matched_rows_target <- input[input$HgncId == edge$target_id & !is.na(input$Protein), ]
+
+    if (nrow(matched_rows_source) != 1 || nrow(matched_rows_target) != 1) {
+        stop(paste0(
+            "INDRA Exception: Unexpected number of matches for the following HGNC IDs in the input data: ", 
+            edge$source_id, 
+            " or ", 
+            edge$target_id, 
+            ". Each ID must match exactly one entry in the input data, but 0 or multiple matches were found. Please check the input data for duplicates or missing entries."
+        ))
+    } 
+
+    edge$source_uniprot_id <- matched_rows_source$Protein
+    edge$target_uniprot_id <- matched_rows_target$Protein
+
     return(edge)
 }
 

tests/testthat/test-getSubnetworkFromIndra.R

@@ -33,7 +33,7 @@ test_that("Exception is thrown for 400+ proteins in dataframe", {
     )
     expect_error(
         getSubnetworkFromIndra(input_400),
-        "Invalid Input Error: INDRA query must contain less than 400 proteins.  Consider adding a p-value cutoff"
+        "Invalid Input Error: INDRA query must contain less than 400 proteins.  Consider lowering your p-value cutoff"
     )
 })