pybiolib 1.2.158.dev1__tar.gz → 1.2.163.dev1__tar.gz

{pybiolib-1.2.158.dev1 → pybiolib-1.2.163.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pybiolib
-Version: 1.2.158.dev1
+Version: 1.2.163.dev1
 Summary: BioLib Python Client
 Home-page: https://github.com/biolib
 License: MIT

{pybiolib-1.2.158.dev1 → pybiolib-1.2.163.dev1}/biolib/utils/seq_util.py

@@ -1,7 +1,7 @@
 import re
 from io import BufferedIOBase, TextIOBase
 
-from biolib.typing_utils import Dict, List, Optional, Union, Iterator
+from biolib.typing_utils import Dict, Iterator, List, Optional, Union
 
 
 class SeqUtilRecord:
@@ -35,14 +35,22 @@ class SeqUtil:
         input_file: Union[str, BufferedIOBase, None] = None,
         default_header: Optional[str] = None,
         allow_any_sequence_characters: bool = False,
+        use_strict_alphabet: Optional[bool] = False,
         allow_empty_sequence: bool = True,
         file_name: Optional[str] = None,
 ) -> Iterator[SeqUtilRecord]:
         def process_and_yield_record(header: str, sequence_lines: List[str]):
             sequence = ''.join(sequence_lines)
             sequence_id = header.split()[0]
+            if allow_any_sequence_characters and use_strict_alphabet:
+                raise Exception(
+                    'Error: Please choose either allow_any_sequence_characters or use_strict_alphabet'
+                )
             if not allow_any_sequence_characters:
-                invalid_sequence_characters = SeqUtil._find_invalid_sequence_characters(sequence)
+                if use_strict_alphabet:
+                    invalid_sequence_characters = SeqUtil._find_invalid_sequence_characters_strict(sequence)
+                else:
+                    invalid_sequence_characters = SeqUtil._find_invalid_sequence_characters(sequence)
                 if invalid_sequence_characters:
                     raise Exception(
                         f'Error: Invalid character ("{invalid_sequence_characters[0]}") found in sequence {sequence_id}'
@@ -126,6 +134,15 @@ class SeqUtil:
         invalid_chars = [char for char in sequence if char not in allowed_sequence_chars]
         return invalid_chars
 
+    @staticmethod
+    def _find_invalid_sequence_characters_strict(sequence: str) -> List[str]:
+        # Equivalent to fair-esm alphabet, compatible with ESM-models
+        # Excludes digits, '_' and 'J' (ambiguous letter only used in mass-spec NMR)
+        # https://github.com/facebookresearch/esm/blob/2b369911bb5b4b0dda914521b9475cad1656b2ac/esm/constants.py#L8
+        allowed_sequence_chars = set('lagvsertidpkqnfymhwcxbuzoLAGVSERTIDPKQNFYMHWCXBUZO-.')
+        invalid_chars = [char for char in sequence if char not in allowed_sequence_chars]
+        return invalid_chars
+
     @staticmethod
     def _find_invalid_sequence_id_characters(sequence: str) -> List[str]:
         allowed_chars = set('abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_.:*#')

{pybiolib-1.2.158.dev1 → pybiolib-1.2.163.dev1}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "pybiolib"
-version = "1.2.158.dev1"
+version = "1.2.163.dev1"
 description = "BioLib Python Client"
 readme = "PYPI_README.md"
 license = "MIT"