feat(#639): add annotated_but_not_staged check

pharme.code-workspace

@@ -35,6 +35,7 @@
       "ABCG",
       "Abilify",
       "abiraterone",
+      "amikacin",
       "anni",
       "aripiprazole",
       "atorvastatin",
@@ -62,6 +63,7 @@
       "Ezallor",
       "fluorouracil",
       "fullscreen",
+      "gentamicin",
       "haplotype",
       "haplotypes",
       "Hasso",
@@ -73,6 +75,7 @@
       "inhibitable",
       "irinotecan",
       "Jantoven",
+      "kanamycin",
       "Keycloak",
       "lookupkey",
       "lookupkeys",
@@ -95,6 +98,7 @@
       "oxcarbazepine",
       "Oxtellar",
       "pantoprazole",
+      "paromomycin",
       "paroxetine",
       "pazopanib",
       "Pharmacogenetic",
@@ -105,6 +109,7 @@
       "PharMe",
       "phenoconversion",
       "Plattner",
+      "plazomicin",
       "Proprinal",
       "pubspec",
       "RGBO",
@@ -122,6 +127,7 @@
       "subfolders",
       "tacrolimus",
       "terbinafine",
+      "tobramycin",
       "tramadol",
       "Trileptal",
       "Ultrarapid",

scripts/README.md

@@ -96,6 +96,7 @@ and optionally correct what can be corrected easily in
 | `brand_whitespace` | Drug brand names should not have leading or trailing white space. | ✅ | ❌ |
 | `single_any_fallback` | If any fallback guidelines `*` are present, only one guideline should be present (otherwise other guidelines are ignored) | ❌ | ❌ |
 | `fallback_single_lookup` | If fallback guidelines `*` or `~` are present, only one lookup value per gene should be present (otherwise other lookup values are ignored) | ❌ | ❌ |
+| `annotated_but_not_staged` | Warns if a drug is annotated but not staged (ignored drugs in `IGNORE_STAGED_CHECK`) | ❌ | ❌ |
 
 ### Guideline annotation checks
 
@@ -108,6 +109,7 @@ and optionally correct what can be corrected easily in
 | `green_warning` | Green warning level should be applied in all non-red and non-yellow cases and when the recommendation states "at standard dose" or similar formulations. | ❌ | ❌ |
 | `none_warning` | None warning level should be applied in all not handled warning level cases. | ❌ | ❌ |
 | `metabolization_before_consequence` | Metabolization implications should come before consequences. | ❌ | ❌ |
+| `annotated_but_not_staged` | Warns if a guideline is annotated but not staged (ignored drugs in `IGNORE_STAGED_CHECK`) | ❌ | ❌ |
 
 \* Skips guidelines with multiple genes unless all results but one are missing
 or indeterminate.

scripts/analyze.py

@@ -1,5 +1,6 @@
 import sys
 
+from analyze_functions.checks.fully_annotated_staged import check_if_fully_annotated_staged
 from analyze_functions.checks.brand_name_whitespace import check_brand_name_whitespace
 from analyze_functions.checks.metabolization_before_consequence import check_metabolization_before_consequence
 from analyze_functions.checks.fallback_guidelines import check_single_any_fallback_guideline, check_single_lookup_fallback_guideline
@@ -21,6 +22,7 @@
     'brand_whitespace': check_brand_name_whitespace,
     'single_any_fallback': check_single_any_fallback_guideline,
     'fallback_single_lookup': check_single_lookup_fallback_guideline,
+    'annotated_but_not_staged': check_if_fully_annotated_staged,
 }
 
 DRUG_CORRECTIONS = {
@@ -35,20 +37,17 @@
     'green_warning_level': check_green_warning_level,
     'none_warning_level': check_none_warning_level,
     'metabolization_before_consequence': check_metabolization_before_consequence,
+    'annotated_but_not_staged': check_if_fully_annotated_staged,
 }
 
 GUIDELINE_CORRECTIONS = {
     'has_consult': add_consult,
 }
 
-def analyze_annotations(item, annotations, checks, data = None):
+def analyze_annotations(checks, check_args):
     results = {}
     for check_name, check_function in checks.items():
-        results[check_name] = check_function({
-            'item': item,
-            'annotations': annotations,
-            'data': data,
-        })
+        results[check_name] = check_function(check_args)
     return results
 
 def correct_inconsistency(data, item, check_name, corrections):
@@ -112,7 +111,14 @@ def run_analyses():
             log_not_annotated(log_content)
         else:
             drug_result = analyze_annotations(
-                drug, drug_annotations, DRUG_CHECKS, data)
+                DRUG_CHECKS,
+                {
+                    'item': drug,
+                    'annotations': drug_annotations,
+                    'data': data,
+                    'drug_name': drug_name,
+                },
+            )
             if not all(drug_result.values()):
                 skipped, failed = handle_failed_checks(data, drug, drug_result,
                     DRUG_CORRECTIONS, correct_inconsistencies,
@@ -131,7 +137,13 @@ def run_analyses():
                 log_not_annotated(log_content)
                 continue
             guideline_result = analyze_annotations(
-                guideline, guideline_annotations, GUIDELINE_CHECKS)
+                GUIDELINE_CHECKS,
+                {
+                    'item': guideline,
+                    'annotations': guideline_annotations,
+                    'drug_name': drug_name,
+                },
+            )
             if guideline_result == None: continue
             if not all(guideline_result.values()):
                 skipped, failed = handle_failed_checks(

scripts/analyze_functions/checks/fully_annotated_staged.py

@@ -0,0 +1,21 @@
+IGNORE_STAGED_CHECK = [#
+    'amikacin',
+    'gentamicin',
+    'kanamycin',
+    'paromomycin',
+    'tobramycin',
+    'streptomycin',
+    'plazomicin',
+]
+
+def check_if_fully_annotated_staged(args):
+    if args['drug_name'] in IGNORE_STAGED_CHECK: return None
+    item = args['item']
+    isStaged = item['isStaged']
+    if isStaged: return True
+    annotations = args['annotations']
+    isFullyAnnotated = all(map(
+        lambda annotationKey: annotations[annotationKey] != None,
+        annotations.keys(),
+    ))
+    return isFullyAnnotated and isStaged