feat(scripts): change how "much" implications are handled

pharme.code-workspace

@@ -82,6 +82,7 @@
       "maxdepth",
       "MedlinePlus",
       "Metabolizer",
+      "metabolizers",
       "mirabegron",
       "mocktail",
       "myopathy",
@@ -122,6 +123,7 @@
       "terbinafine",
       "tramadol",
       "Trileptal",
+      "Ultrarapid",
       "Unmappable",
       "unscrollable",
       "unstage",

scripts/README.md

@@ -92,7 +92,7 @@ optionally correct what can be corrected easily in
 | Check | Description | `--correct`ed | Only for single-gene results* |
 | ----- | ----------- | ------------- | ----------------------------- |
 | `has_consult` | Is "consult your pharmacist..." included in recommendation? | ✅ | ❌ |
-| `implication_severity` | Poor/ultrarapid phenotypes with "faster" or "slower" implication should have "much" keyword, intermediate/rapid not. | ❌ | ✅ |
+| `implication_severity` | "Much" keyword, should only be used if reflected by guideline implication. | ❌ | ✅ |
 | `red_warning` | Red warning level should always have recommendation "may not be the right medication" and vice versa. | ❌ | ❌ |
 | `yellow_warning` | Recommendation containing "adjusted" or "higher" or "lower" but not "may not be the right" should have yellow warning level. | ❌ | ❌ |
 | `green_warning` | Green warning level should have recommendation "at standard dose" but not "adjusted" and vise versa. | ❌ | ❌ |

scripts/analyze.py

@@ -66,29 +66,52 @@ def has_consult(_, annotations):
     return CONSULT_TEXT in annotations['recommendation']
 
 def check_implication_severity(guideline, annotations):
-    phenotype = get_phenotype_key(guideline).lower()
-    gene_number = len(guideline['phenotypes'].keys())
-    missing_genes = phenotype.count('no result') + \
-        phenotype.count('indeterminate')
-    if gene_number - missing_genes != 1:
+    ignored_phenotypes = ['no result', 'indeterminate', 'normal metabolizer']
+    multiple_relevant_phenotypes = False
+    relevant_gene = None
+    for current_gene, current_phenotypes in guideline['phenotypes'].items():
+        if not current_phenotypes[0].lower() in ignored_phenotypes:
+            if relevant_gene != None:
+                multiple_relevant_phenotypes = True
+                break
+            relevant_gene = current_gene
+    if multiple_relevant_phenotypes or relevant_gene == None:
         return None
-    severity_rules = [
-        { 'has_much': True, 'phenotype': 'ultrarapid', 'implication': 'faster' },
-        { 'has_much': True, 'phenotype': 'poor', 'implication': 'slower' },
-        { 'has_much': False, 'phenotype': 'rapid', 'implication': 'faster' },
-        { 'has_much': False, 'phenotype': 'intermediate', 'implication': 'slower' },
+    implication = \
+        guideline['externalData'][0]['implications'][relevant_gene].lower()
+    much_implying_formulations = [
+        'greatly decreased',
+        'greatly reduced',
+        'significantly reduced',
+        'extremely high concentrations',
+        'when compared to cyp2c19 rapid and normal metabolizers',
+        'as compared to non-poor metabolizers',
+        'when compared to cyp2c19 normal and intermediate metabolizers',
+        'as compared to normal and intermediate metabolizer',
+        'complete dpd deficiency',
     ]
-    check_applies = True
-    for severity_rule in severity_rules:
-        if severity_rule['phenotype'] == 'rapid' and 'ultrarapid' in phenotype:
-            continue
-        rule_broken = severity_rule['phenotype'] in phenotype and \
-            severity_rule['implication'] in annotations['implication'] and \
-                severity_rule['has_much'] != ('much' in annotations['implication'])
-        if rule_broken:
-            check_applies = False
-            break
-    return check_applies
+    much_formulations = [
+        'much faster',
+        'much slower'
+    ]
+    much_is_implied = any(
+        map(
+            lambda much_implying_formulation:
+                much_implying_formulation in implication,
+            much_implying_formulations,
+        )
+    )
+    implication_has_much = any(
+        map(
+            lambda much_formulation: much_formulation in annotations['implication'],
+            much_formulations,
+        )
+    )
+    if not much_is_implied == implication_has_much:
+        print(implication_has_much)
+        print(annotations['implication'])
+        print(implication)
+    return much_is_implied == implication_has_much
 
 def check_red_warning_level(_, annotations):
     has_warning_level = annotations['warning_level'] == 'red'