FInd similar alleles/MACs

CHANGELOG.md

@@ -1,3 +1,52 @@
+<a name="1.0.6"></a>
+# [1.0.6 Validation of allele specific MAC codes](https://github.com/nmdp-bioinformatics/py-ard/releases/tag/1.0.6) - 30 Oct 2023
+
+- Use allele specific antigen code rules when validating MACs that cross antigen group similar to [MAC Service](https://hml.nmdp.org/macui/)
+- Returns the original `InvalidAlleleError` instead of wrapping it in `InvalidTypingError` when an allele is not valid.
+
+[Changes][1.0.6]
+
+
+<a name="1.0.5"></a>
+# [1.0.5 Non strict mode](https://github.com/nmdp-bioinformatics/py-ard/releases/tag/1.0.5) - 04 Oct 2023
+
+Supports non-strict mode makes valid alleles by adding expression characters to invalid alleles.
+
+## Use non `strict` mode in config to reduce alleles that may be valid with expression characters.
+
+```python
+>>> my_configs = {'strict': False, 'verbose_log': True}
+>>> import pyard
+>>> ard = pyard.init(config=my_configs, load_mac=False)
+
+>>> ard.redux('A*24:329', 'lgx')
+A*24:329 is not valid. Using A*24:329Q
+'A*24:329Q'
+
+>>> ard.redux('DQB1*03:276', 'lgx')
+DQB1*03:276 is not valid. Using DQB1*03:276N
+'DQB1*03:01'
+```
+
+## Add non-strict and verbose modes to pyard CLI.
+
+```bash
+❯ pyard --gl "DQB1*03:276" -r lgx
+Typing Error: DQB1*03:276 is not valid GL String.
+ DQB1*03:276 is not a valid Allele
+
+❯ pyard --non-strict --gl "DQB1*03:276" -r lgx
+DQB1*03:01
+
+❯ pyard --non-strict --verbose --gl "DQB1*03:276" -r lgx
+DQB1*03:276 is not valid. Using DQB1*03:276N
+DQB1*03:01
+```
+
+
+[Changes][1.0.5]
+
+
 <a name="1.0.4"></a>
 # [Fixes when used without login user (1.0.4)](https://github.com/nmdp-bioinformatics/py-ard/releases/tag/1.0.4) - 19 Sep 2023
 
@@ -563,6 +612,8 @@ yes
 [Changes][0.0.14]
 
 
+[1.0.6]: https://github.com/nmdp-bioinformatics/py-ard/compare/1.0.5...1.0.6
+[1.0.5]: https://github.com/nmdp-bioinformatics/py-ard/compare/1.0.4...1.0.5
 [1.0.4]: https://github.com/nmdp-bioinformatics/py-ard/compare/1.0.3...1.0.4
 [1.0.3]: https://github.com/nmdp-bioinformatics/py-ard/compare/1.0.2...1.0.3
 [1.0.2]: https://github.com/nmdp-bioinformatics/py-ard/compare/1.0.1...1.0.2

Dockerfile

@@ -4,7 +4,7 @@ LABEL MAINTAINER="Pradeep Bashyal"
 
 WORKDIR /app
 
-ARG PY_ARD_VERSION=1.0.5
+ARG PY_ARD_VERSION=1.0.6
 
 COPY requirements.txt /app
 RUN pip install --no-cache-dir --upgrade pip && \

api-spec.yaml

@@ -2,7 +2,7 @@ openapi: 3.0.3
 info:
   title: ARD Reduction
   description: Reduce to ARD Level
-  version: "1.0.5"
+  version: "1.0.6"
 servers:
   - url: 'http://localhost:8080'
 tags:

pyard/__init__.py

@@ -27,7 +27,7 @@
 from .misc import get_imgt_db_versions as db_versions
 
 __author__ = """NMDP Bioinformatics"""
-__version__ = "1.0.5"
+__version__ = "1.0.6"
 
 
 def init(

pyard/ard.py

@@ -25,6 +25,7 @@
 import re
 import sqlite3
 import sys
+from collections import Counter
 from typing import Iterable, List
 
 from . import broad_splits, smart_sort
@@ -408,8 +409,8 @@ def redux(self, glstring: str, redux_type: VALID_REDUCTION_TYPES) -> str:
                     )
 
         # Handle MAC
-        if self._config["reduce_MAC"] and self.is_mac(glstring):
-            if db.is_valid_mac_code(self.db_connection, code):
+        if self._config["reduce_MAC"] and code.isalpha():
+            if self.is_mac(glstring):  # Make sure it's a valid MAC
                 if HLA_regex.search(glstring):
                     # Remove HLA- prefix
                     allele_name = glstring.split("-")[1]
@@ -436,12 +437,7 @@ def validate(self, glstring):
         :param glstring: GL String to validate
         :return: boolean indicating success
         """
-        try:
-            return self._is_valid_gl(glstring)
-        except InvalidAlleleError as e:
-            raise InvalidTypingError(
-                f"{glstring} is not valid GL String. \n {e.message}", e
-            ) from None
+        return self._is_valid_gl(glstring)
 
     def is_XX(self, glstring: str, loc_antigen: str = None, code: str = None) -> bool:
         if loc_antigen is None or code is None:
@@ -484,12 +480,39 @@ def is_mac(self, allele: str) -> bool:
         :return: True if MAC
         """
         if ":" in allele:
-            code = allele.split(":")[1]
-            try:
+            allele_split = allele.split(":")
+            if len(allele_split) == 2:  # MACs have only single :
+                locus_antigen, code = allele_split
                 if code.isalpha():
-                    return db.is_valid_mac_code(self.db_connection, code)
-            except sqlite3.OperationalError as e:
-                print("Error: ", e)
+                    try:
+                        alleles = db.mac_code_to_alleles(self.db_connection, code)
+                        if alleles:
+                            if any(map(lambda a: ":" in a, alleles)):
+                                # allele specific antigen codes have ':' in the MAC mapping
+                                # e.g. CFWRN -> 15:01/15:98/15:157/15:202/
+                                #               15:239/15:280/15:340/35:43/35:67/35:79/35:102/35:118/35:185/51:220
+                                # Extract the antigens from the mapped alleles
+                                antigen_groups = map(lambda a: a.split(":")[0], alleles)
+                                # Rule 1: The 1st field with the most allele designations in the request is
+                                #         the 1st field of the allele code designation
+                                # Rule 2: If there is a tie in the number of alleles designations sharing the 1st field,
+                                #         the 1st field with the lowest numeric value is selected.
+                                antigen_counts = Counter(antigen_groups)
+                                # Create a table of antigen to it's counts
+                                # '15': 7
+                                # '35': 6
+                                # '51': 1
+                                # Valid antigen is the first most common one.
+                                # As it's presorted in db, also satisfies Rule 2.
+                                valid_antigen = antigen_counts.most_common(1).pop()[0]
+                                # Get antigen value 15 from 'DRB1*15'
+                                provided_antigen = locus_antigen.split("*").pop()
+                                # The MAC is only valid if the given antigen satisfies the antigen matching Rule 1 and 2
+                                return provided_antigen == valid_antigen
+                            # Valid when antigen group codes
+                            return True
+                    except sqlite3.OperationalError as e:
+                        print("Error: ", e)
         return False
 
     def is_v2(self, allele: str) -> bool:
@@ -719,8 +742,8 @@ def expand_mac(self, mac_code: str):
         :return: GL String of expanded alleles
         :rtype: str
         """
-        locus_antigen, code = mac_code.split(":")
-        if db.is_valid_mac_code(self.db_connection, code):
+        if self.is_mac(mac_code):  # Validate MAC first
+            locus_antigen, code = mac_code.split(":")
             if HLA_regex.search(mac_code):
                 locus_antigen = locus_antigen.split("-")[1]  # Remove HLA- prefix
                 return "/".join(

pyard/db.py

@@ -172,21 +172,6 @@ def alleles_to_mac_code(
     return None
 
 
-def is_valid_mac_code(connection: sqlite3.Connection, code: str) -> bool:
-    """
-    Check db if the MAC code exists.
-
-    :param connection: db connection of type sqlite.Connection
-    :param code: MAC code
-    :return: code is MAC code ?
-    """
-    mac_query = "SELECT count(alleles) from mac_codes where code = ?"
-    cursor = connection.execute(mac_query, (code,))
-    result = cursor.fetchone()
-    cursor.close()
-    return result[0] > 0
-
-
 def serology_to_alleles(connection: sqlite3.Connection, serology: str) -> List[str]:
     """
     Look up Serology in the database and return corresponding list of alleles.
@@ -371,16 +356,16 @@ def load_dict(
     return table_as_dict
 
 
-def similar_alleles(connection: sqlite3.Connection, allele_name: str) -> Set[str]:
+def similar_alleles(connection: sqlite3.Connection, allele_prefix: str) -> Set[str]:
     """
-    Find similar alleles starting with the provided allele_name.
+    Find similar alleles starting with the provided prefix.
 
     :param connection: db connection of type sqlite.Connection
-    :param allele_name: Allele name to use as a prefix to find similar alleles
+    :param allele_prefix: Allele name to use as a prefix to find similar alleles
     :return: list of similar alleles
     """
     query = "SELECT allele FROM alleles WHERE allele LIKE ?"
-    cursor = connection.execute(query, (f"{allele_name}%",))
+    cursor = connection.execute(query, (f"{allele_prefix}%",))
     result = cursor.fetchall()
     # fetchall() returns a list of tuples of results
     # e.g. [('C*04:09N',)]
@@ -389,6 +374,24 @@ def similar_alleles(connection: sqlite3.Connection, allele_name: str) -> Set[str
     return alleles
 
 
+def similar_mac(connection: sqlite3.Connection, mac_prefix: str) -> Set[str]:
+    """
+    Find similar MAC codes starting with the provided prefix.
+
+    :param connection: db connection of type sqlite.Connection
+    :param mac_prefix: MAC fragment to use as a prefix to find similar MACs
+    :return: list of similar MAC codes
+    """
+    query = "SELECT code FROM mac_codes WHERE code LIKE ?"
+    cursor = connection.execute(query, (f"{mac_prefix}%",))
+    result = cursor.fetchall()
+    # fetchall() returns a list of tuples of results
+    # e.g. [('DJZUP',)]
+    # Get out the first value of the tuple from the result list
+    codes = set(map(lambda t: t[0], result))
+    return codes
+
+
 def find_serology_for_allele(
     connection: sqlite3.Connection, allele_name: str
 ) -> Dict[str, str]: