Merge pull request #310 from pbashyal-nmdp/serology_to_xx

mmaiers-nmdp · web-flow · commit 2fbe8fabf089 · 2024-03-04T17:33:05.000-06:00
Add XX mappings of Serology
diff --git a/pyard/ard.py b/pyard/ard.py
@@ -124,7 +124,7 @@ def __init__(
         dr.generate_serology_mapping(
             self.db_connection, imgt_version, self.serology_mapping, self._redux_allele
         )
-        self.valid_serology_set = dr.build_valid_serology_set(self.db_connection)
+        self.valid_serology_set = SerologyMapping.get_valid_serology_names()
 
         # Load V2 to V3 mappings
         dr.generate_v2_to_v3_mapping(self.db_connection, imgt_version)
@@ -436,16 +436,23 @@ def redux(self, glstring: str, redux_type: VALID_REDUCTION_TYPES) -> str:
             is_hla_prefix = HLA_regex.search(loc_antigen)
             if is_hla_prefix:
                 loc_antigen = loc_antigen.split("-")[1]
-            if self.is_XX(glstring, loc_antigen, code):
-                if is_hla_prefix:
-                    reduced_alleles = self.redux(
-                        "/".join(self.code_mappings.xx_codes[loc_antigen]), redux_type
-                    )
-                    return "/".join(["HLA-" + a for a in reduced_alleles.split("/")])
+            if code == "XX":
+                if self.is_XX(glstring, loc_antigen, code):
+                    if is_hla_prefix:
+                        reduced_alleles = self.redux(
+                            "/".join(self.code_mappings.xx_codes[loc_antigen]),
+                            redux_type,
+                        )
+                        return "/".join(
+                            ["HLA-" + a for a in reduced_alleles.split("/")]
+                        )
+                    else:
+                        return self.redux(
+                            "/".join(self.code_mappings.xx_codes[loc_antigen]),
+                            redux_type,
+                        )
                 else:
-                    return self.redux(
-                        "/".join(self.code_mappings.xx_codes[loc_antigen]), redux_type
-                    )
+                    raise InvalidTypingError(f"{glstring} is not valid XX code")
 
         # Handle MAC
         if self._config["reduce_MAC"] and code.isalpha():
@@ -633,7 +640,13 @@ def find_broad_splits(self, allele) -> tuple:
         return self.serology_mapping.find_splits(allele)
 
     def find_associated_antigen(self, serology) -> str:
-        return self.serology_mapping.serology_associated_map.get(serology, serology)
+        return self.serology_mapping.find_associated_antigen(serology)
+
+    @functools.lru_cache()
+    def find_xx_from_serology(self, serology):
+        if self.is_serology(serology):
+            return db.find_xx_for_serology(self.db_connection, serology)
+        raise InvalidAlleleError(f"{serology} is not a valid serology")
 
     def _get_alleles(self, code, locus_antigen) -> Iterable[str]:
         """
diff --git a/pyard/data_repository.py b/pyard/data_repository.py
@@ -27,15 +27,14 @@
 import pyard.load
 from pyard.smart_sort import smart_sort_comparator
 from . import db
-from .serology import broad_splits_dna_mapping, get_all_valid_serology_names
+from .constants import expression_chars
 from .load import (
     load_g_group,
     load_p_group,
     load_allele_list,
     load_serology_mappings,
     load_latest_version,
 )
-from .constants import expression_chars
 from .mappings import (
     ars_mapping_tables,
     ARSMapping,
@@ -50,6 +49,7 @@
     number_of_fields,
     get_1field_allele,
 )
+from .serology import broad_splits_dna_mapping, SerologyMapping
 
 
 def expression_reduce(df):
@@ -356,7 +356,10 @@ def to_serological_name(locus_name: str):
 
 
 def generate_serology_mapping(
-    db_connection: sqlite3.Connection, imgt_version, serology_mapping, redux_function
+    db_connection: sqlite3.Connection,
+    imgt_version: str,
+    serology_mapping: SerologyMapping,
+    redux_function,
 ):
     if not db.table_exists(db_connection, "serology_mapping"):
         df_sero = load_serology_mappings(imgt_version)
@@ -412,22 +415,28 @@ def generate_serology_mapping(
                     if split in sero_mapping:
                         sero_mapping[broad] = sero_mapping[split]
 
-        # re-sort allele lists into smartsort order
-        for sero in sero_mapping.keys():
-            sero_mapping[sero] = (
-                "/".join(
-                    sorted(
-                        sero_mapping[sero][0],
-                        key=functools.cmp_to_key(smart_sort_comparator),
-                    )
-                ),
-                "/".join(
-                    sorted(
-                        sero_mapping[sero][1],
-                        key=functools.cmp_to_key(smart_sort_comparator),
+        # Create a mapping of serology to alleles, lgx_alleles and associated XX allele
+        serology_xx_mapping = serology_mapping.get_xx_mappings()
+        # re-sort allele lists into smart-sort order
+        for sero in serology_xx_mapping:
+            if sero in sero_mapping:
+                sero_mapping[sero] = (
+                    "/".join(
+                        sorted(
+                            sero_mapping[sero][0],
+                            key=functools.cmp_to_key(smart_sort_comparator),
+                        )
+                    ),
+                    "/".join(
+                        sorted(
+                            sero_mapping[sero][1],
+                            key=functools.cmp_to_key(smart_sort_comparator),
+                        ),
                     ),
-                ),
-            )
+                    serology_xx_mapping[sero],
+                )
+            else:
+                sero_mapping[sero] = (None, None, serology_xx_mapping[sero])
 
         db.save_serology_mappings(db_connection, sero_mapping)
 
@@ -483,12 +492,3 @@ def generate_cwd_mapping(db_connection: sqlite3.Connection):
     if not db.table_exists(db_connection, "cwd2"):
         cwd2_map = pyard.load.load_cwd2()
         db.save_cwd2(db_connection, cwd2_map)
-
-
-def build_valid_serology_set(db_connection: sqlite3.Connection):
-    valid_serology_names = get_all_valid_serology_names()
-    # Save to db if `valid_serology` table is not present
-    if not db.table_exists(db_connection, "valid_serology"):
-        db.save_set(db_connection, "valid_serology", valid_serology_names, "serology")
-
-    return set(valid_serology_names)
diff --git a/pyard/db.py b/pyard/db.py
@@ -412,6 +412,22 @@ def find_serology_for_allele(
     return serology_mapping
 
 
+def find_xx_for_serology(connection: sqlite3.Connection, serology: str) -> str:
+    """
+    Find the corresponding XX allele for the given serology
+
+    :param connection: db connection of type sqlite.Connection
+    :param serology: serology for which to find XX allele
+    :return:  XX allele for given serology
+    """
+    query = f"SELECT xx FROM serology_mapping WHERE serology = ?"
+    cursor = connection.execute(query, (serology,))
+    results = cursor.fetchone()
+    if results:
+        return results[0]
+    return None
+
+
 def get_user_version(connection: sqlite3.Connection) -> int:
     """
     Retrieve user_version from db
@@ -424,9 +440,7 @@ def get_user_version(connection: sqlite3.Connection) -> int:
     version = result[0]
     cursor.close()
 
-    if version:
-        return version
-    return None
+    return version
 
 
 def set_user_version(connection: sqlite3.Connection, version: int):
@@ -579,15 +593,16 @@ def save_serology_mappings(db_connection, sero_mapping):
     # Create table
     create_table_sql = f"""CREATE TABLE serology_mapping (
                             serology TEXT PRIMARY KEY,
-                            allele_list TEXT NOT NULL,
-                            lgx_allele_list TEXT NOT NULL
+                            allele_list TEXT,
+                            lgx_allele_list TEXT,
+                            xx TEXT NOT NULL
                     )"""
     cursor.execute(create_table_sql)
 
-    rows = ((k, v[0], v[1]) for k, v in sero_mapping.items())
+    rows = ((k, v[0], v[1], v[2]) for k, v in sero_mapping.items())
 
     # insert
-    cursor.executemany(f"INSERT INTO serology_mapping VALUES (?, ?, ?)", rows)
+    cursor.executemany(f"INSERT INTO serology_mapping VALUES (?, ?, ?, ?)", rows)
 
     # commit transaction - writes to the db
     db_connection.commit()
diff --git a/pyard/serology.py b/pyard/serology.py
@@ -20,15 +20,72 @@
 #    > http://www.fsf.org/licensing/licenses/lgpl.html
 #    > http://www.opensource.org/licenses/lgpl-license.php
 #
-from pyard.constants import HLA_regex
+import re
 
+from pyard.constants import HLA_regex
 
 #
 # HLA Antigens
 # List of all recognised serological collected from:
 # https://hla.alleles.org/antigens/recognised_serology.html
 #
-def get_all_valid_serology_names():
+
+
+# -#
+# Broad, Splits and Associated Antigens
+# http://hla.alleles.org/antigens/broads_splits.html
+#
+#
+# Mapping Generated from `dna_relshp.csv` file
+#
+broad_splits_dna_mapping = {
+    "A*09": ["A*23", "A*24"],
+    "A*10": ["A*25", "A*26", "A*34", "A*66"],
+    "A*19": ["A*29", "A*30", "A*31", "A*32", "A*33", "A*74"],
+    "A*28": ["A*68", "A*69"],
+    "B*05": ["B*51", "B*52"],
+    "B*12": ["B*44", "B*45"],
+    "B*16": ["B*38", "B*39"],
+    "B*17": ["B*57", "B*58"],
+    "B*21": ["B*49", "B*50"],
+    "B*22": ["B*54", "B*55", "B*56"],
+    "C*10": ["C*03", "C*04"],
+    "DQB1*01": ["DQB1*05", "DQB1*06"],
+    "DRB1*02": ["DRB1*15", "DRB1*16"],
+    "DRB1*06": ["DRB1*13", "DRB1*14"],
+}
+
+serology_xx_exception_mapping = {
+    # Locus B
+    # Broad B40
+    "B60": "B*40:XX",
+    "B61": "B*40:XX",
+    # Broad B14
+    "B64": "B*14:XX",
+    "B65": "B*14:XX",
+    # Broad B15
+    "B62": "B*15:XX",
+    "B63": "B*15:XX",
+    "B70": "B*15:XX",
+    "B75": "B*15:XX",
+    "B76": "B*15:XX",
+    "B77": "B*15:XX",
+    # Broad B70
+    "B71": "B*15:XX",
+    "B72": "B*15:XX",
+    "DR17": "DRB1*03:XX",
+    "DR18": "DRB1*03:XX",
+    # Locus DQB1
+    # Broad DQ3
+    "DQ7": "DQB1*03:XX",
+    "DQ8": "DQB1*03:XX",
+    "DQ9": "DQB1*03:XX",
+}
+
+sero_antigen_regex = re.compile(r"(\D+)(\d+)")
+
+
+class SerologyMapping:
     valid_serology_map = {
         "A": [
             "A1",
@@ -153,7 +210,7 @@ def get_all_valid_serology_names():
             "Dw25",
             "Dw26",
         ],
-        "DR": [
+        "DRB1": [
             "DR1",
             "DR103",
             "DR2",
@@ -179,40 +236,10 @@ def get_all_valid_serology_names():
             "DR52",
             "DR53",
         ],
-        "DQ": ["DQ1", "DQ2", "DQ3", "DQ4", "DQ5", "DQ6", "DQ7", "DQ8", "DQ9"],
-        "DP": ["DPw1", "DPw2", "DPw3", "DPw4", "DPw5", "DPw6"],
+        "DQB1": ["DQ1", "DQ2", "DQ3", "DQ4", "DQ5", "DQ6", "DQ7", "DQ8", "DQ9"],
+        "DPB1": ["DPw1", "DPw2", "DPw3", "DPw4", "DPw5", "DPw6"],
     }
 
-    all_serology_names = [x for v in valid_serology_map.values() for x in v]
-    return all_serology_names
-
-
-# -#
-# Broad, Splits and Associated Antigens
-# http://hla.alleles.org/antigens/broads_splits.html
-#
-#
-# Mapping Generated from `dna_relshp.csv` file
-#
-broad_splits_dna_mapping = {
-    "A*09": ["A*23", "A*24"],
-    "A*10": ["A*25", "A*26", "A*34", "A*66"],
-    "A*19": ["A*29", "A*30", "A*31", "A*32", "A*33", "A*74"],
-    "A*28": ["A*68", "A*69"],
-    "B*05": ["B*51", "B*52"],
-    "B*12": ["B*44", "B*45"],
-    "B*16": ["B*38", "B*39"],
-    "B*17": ["B*57", "B*58"],
-    "B*21": ["B*49", "B*50"],
-    "B*22": ["B*54", "B*55", "B*56"],
-    "C*10": ["C*03", "C*04"],
-    "DQB1*01": ["DQB1*05", "DQB1*06"],
-    "DRB1*02": ["DRB1*15", "DRB1*16"],
-    "DRB1*06": ["DRB1*13", "DRB1*14"],
-}
-
-
-class SerologyMapping:
     def __init__(self, broad_splits_mapping, associated_mapping):
         self.broad_splits_map = broad_splits_mapping
         self.serology_associated_map = associated_mapping
@@ -237,8 +264,43 @@ def find_splits(self, allele: str) -> tuple:
             if allele_name in mapping[broad]:
                 return self._get_mapping(broad, mapping, prefix)
 
-    @staticmethod
-    def _get_mapping(broad, mapping, prefix):
+    def find_associated_antigen(self, serology):
+        return self.serology_associated_map.get(serology, serology)
+
+    def get_xx_mappings(self):
+        all_xx_mappings = {}
+        for locus, serologies in SerologyMapping.valid_serology_map.items():
+            xx_mapping = {
+                serology: self._map_serology_to_xx(locus, serology)
+                for serology in serologies
+            }
+            all_xx_mappings.update(xx_mapping)
+        return all_xx_mappings
+
+    @classmethod
+    def get_valid_serology_names(cls):
+        all_serology_names = {x for v in cls.valid_serology_map.values() for x in v}
+        return all_serology_names
+
+    def _map_serology_to_xx(self, locus, serology):
+        if serology in serology_xx_exception_mapping.keys():
+            return serology_xx_exception_mapping[serology]
+
+        # Use the associated serology for XX version
+        serology = self.find_associated_antigen(serology)
+
+        # Extract just the digits
+        antigen_group = sero_antigen_regex.match(serology).group(2)
+        # Pad numbers with 0 for single digit numbers
+        antigen_group_num = int(antigen_group)
+        if antigen_group_num < 10:
+            antigen_group = f"{antigen_group_num:02}"
+
+        # Build the XX allele
+        return f"{locus}*{antigen_group}:XX"
+
+    @classmethod
+    def _get_mapping(cls, broad, mapping, prefix):
         if prefix:
             return "HLA-" + broad, list(map(lambda x: "HLA-" + x, mapping[broad]))
         else:
diff --git a/tests/features/serology.feature b/tests/features/serology.feature
diff --git a/tests/steps/redux_allele.py b/tests/steps/redux_allele.py
