Merge pull request #68 from pbashyal-nmdp/predict_v3_from_v2

Predict v3 from v2

Author: mmaiers-nmdp

README.rst

@@ -11,7 +11,7 @@ py-ard
         :alt: Documentation Status
 
 
-ARD reduction for HLA with python
+ARD reduction for HLA with Python
 
 * Free software: LGPL 3.0
 * Documentation: https://pyard.readthedocs.io.
@@ -68,26 +68,33 @@ Example
     # You can choose to refresh the MAC code for previously used versions
     # ard =  pyard.ARD(3290, refresh_mac=True)
 
-    # Allele to reduce
+    #
+    # Reduce Allele
+    #
     allele = "A*01:01:01"
 
     ard.redux(allele, 'G')
-    # 'A*01:01:01G'
+    # >>> 'A*01:01:01G'
 
     ard.redux(allele, 'lg')
-    # 'A*01:01g'
+    # >>> 'A*01:01g'
 
     ard.redux(allele, 'lgx')
-    # 'A*01:01'
+    # >>> 'A*01:01'
 
+    #
+    # Reduce GL String
+    #
     ard.redux_gl("A*01:01/A*01:01N+A*02:AB^B*07:02+B*07:AB", "G")
     # 'B*07:02:01G+B*07:02:01G^A*01:01:01G+A*02:01:01G/A*02:02'
 
     # py-ard can also reduce serology based typings
     ard.redux_gl('HLA-A*10^HLA-A*9', 'lg')
+    # >>> ard_gl
     # 'HLA-A*24:19g/HLA-A*24:22g^HLA-A*26:01g/HLA-A*26:10g/HLA-A*26:15g/HLA-A*26:92g/HLA-A*66:01g/HLA-A*66:03g'
 
 
+
 Command Line Tools
 ------------------
 
@@ -107,6 +114,9 @@ Command Line Tools
     Created py-ard version 3290 database
     Updated v2_mapping table with 'map2to3.csv' mapping file.
 
+    # Replace the Latest IMGT database with V2 mappings
+    $ pyard-import  --v2-to-v3-mapping map2to3.csv
+
     # Reduce a gl string from command line
     $ pyard --gl 'A*01:AB' -r lgx
     A*01:01/A*01:02
@@ -115,4 +125,4 @@ Command Line Tools
     DRB1*08:01:01G/DRB1*08:02:01G/DRB1*08:03:02G/DRB1*08:04:01G/DRB1*08:05/ ...
 
     $ pyard -v 3290 --gl 'A1' -r lgx
-    A*01:01/A*01:02/A*01:03/A*01:06/A*01:07/A*01:08/A*01:09/A*01:10/A*01:12/ ...
+    A*01:01/A*01:02/A*01:03/A*01:06/A*01:07/A*01:08/A*01:09/A*01:10/A*01:12/ ...

pyard/pyard.py

@@ -188,7 +188,7 @@ def redux_gl(self, glstring: str, redux_type: str) -> str:
         loc_antigen, code = loc_allele[0], loc_allele[1]
 
         # Handle XX codes
-        if self.is_mac(glstring) and code == "XX" and loc_antigen in self.xx_codes:
+        if self.is_XX(glstring, loc_antigen, code):
             return self.redux_gl("/".join(self.xx_codes[loc_antigen]), redux_type)
 
         # Handle MAC
@@ -205,6 +205,12 @@ def redux_gl(self, glstring: str, redux_type: str) -> str:
 
         return self.redux(glstring, redux_type)
 
+    def is_XX(self, glstring: str, loc_antigen: str = None, code: str = None) -> bool:
+        if loc_antigen is None or code is None:
+            loc_allele = glstring.split(":")
+            loc_antigen, code = loc_allele[0], loc_allele[1]
+        return self.is_mac(glstring) and code == "XX" and loc_antigen in self.xx_codes
+
     @staticmethod
     def is_serology(allele: str) -> bool:
         """
@@ -248,7 +254,7 @@ def is_v2(allele: str) -> bool:
         :param allele: Possible allele
         :return: Is the allele in V2 nomenclature
         """
-        return '*' in allele and not ':' in allele
+        return '*' in allele and ':' not in allele
 
     def _is_valid_allele(self, allele):
         """
@@ -290,13 +296,53 @@ def _get_alleles_from_serology(self, serology) -> Iterable[str]:
         else:
             return alleles
 
+    def _combine_with_colon(self, digits_field):
+        num_of_digits = len(digits_field)
+        return ':'.join(digits_field[i:i + 2] for i in range(0, num_of_digits, 2))
+
+    def _predict_v3(self, v2_allele: str) -> str:
+        """
+        Use heuristic to predict V3 from V2
+
+        :param v2_allele: Allele in V2 format
+        :return: V3 format of V2 allele
+        """
+        # Separate out the locus and the allele name part
+        locus, allele_name = v2_allele.split('*')
+        # Separate out the numeric and non-numeric components
+        components = re.findall(r'^(\d+)(.*)', allele_name)
+        if not components:
+            return v2_allele
+        digits_field, non_digits_field = components.pop()
+        # final_allele is the result of the transformation
+        final_allele = digits_field
+        num_of_digits = len(digits_field)
+        if num_of_digits == 1:
+            return v2_allele
+        if num_of_digits > 2:
+            if locus.startswith('DP') and num_of_digits == 5:  # covers DPs with 5 digits
+                final_allele = digits_field[:3] + ':' + (digits_field[3:]) + non_digits_field
+            elif num_of_digits % 2 == 0:  # covers digits with 2, 4, 6, 8
+                final_allele = self._combine_with_colon(digits_field) + non_digits_field
+            else:
+                final_allele = digits_field[:2] + ':' + (digits_field[2:]) + non_digits_field
+        else:
+            if non_digits_field:
+                final_allele = digits_field + ':' + non_digits_field
+        return locus + '*' + final_allele
+
     def _map_v2_to_v3(self, v2_allele):
         """
         Get V3 version of V2 versioned allele
         :param v2_allele: V2 versioned allele
         :return: V3 versioned allele
         """
-        return v2_to_v3_allele(self.db_connection, v2_allele)
+        # Check if it's in the exception case mapping
+        v3_allele = v2_to_v3_allele(self.db_connection, v2_allele)
+        if not v3_allele:
+            # Try and predict V3
+            v3_allele = self._predict_v3(v2_allele)
+        return v3_allele
 
     def isvalid(self, allele: str) -> bool:
         """
@@ -404,3 +450,14 @@ def expand_mac(self, mac_code: str):
                 return list(self._get_alleles(code, locus_antigen))
 
         return ''
+
+    def v2_to_v3(self, v2_allele) -> str:
+        """
+        Convert Version 2 Allele Name to Version 3 Allele Name
+
+        :param v2_allele: Version 2 Allele Name
+        :return: Version 3 Allele Name
+        """
+        if self.is_v2(v2_allele):
+            return self._map_v2_to_v3(v2_allele)
+        return v2_allele

scripts/pyard-import

@@ -92,11 +92,12 @@ if __name__ == '__main__':
         print(f"Created py-ard version {imgt_version} database")
     else:
         ard = pyard.ARD(data_dir=data_dir)
+        imgt_version = 'Latest'
         print(f"Created Latest py-ard database")
     del ard
 
     if v2_to_v3_dict:
         db_connection = db.create_db_connection(data_dir, imgt_version, ro=False)
         db.save_dict(db_connection, table_name='v2_mapping',
                      dictionary=v2_to_v3_dict, columns=('v2', 'v3'))
-        print(f"Updated v2_mapping table with '{args.v2_v3_mapping}' mapping file.")
+        print(f"Updated v2_mapping table with '{args.v2_v3_mapping}' mapping file for {imgt_version} IMGT database.")