[ENH] pdb to String loader using SEQREQ records

examples/aptanet_tutorial.ipynb

@@ -40,7 +40,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": null,
    "id": "3737da88",
    "metadata": {},
    "outputs": [],
@@ -49,7 +49,7 @@
     "import torch\n",
     "\n",
     "from pyaptamer.datasets import load_1gnh_structure\n",
-    "from pyaptamer.utils.struct_to_aaseq import struct_to_aaseq"
+    "from pyaptamer.utils import struct_to_aaseq"
    ]
   },
   {

pyaptamer/datasets/__init__.py

@@ -2,7 +2,7 @@
 
 from pyaptamer.datasets._loaders._one_gnh import load_1gnh_structure
 from pyaptamer.datasets._loaders._online_databank import load_from_rcsb
-from pyaptamer.datasets._loaders._pfoa_loader import load_pfoa_structure
+from pyaptamer.datasets._loaders._pfoa import load_pfoa_structure
 
 __all__ = [
     "load_pfoa_structure",

pyaptamer/datasets/_loaders/__init__.py

@@ -1,6 +1,6 @@
 """Loaders for different data structures."""
 
 from pyaptamer.datasets._loaders._one_gnh import load_1gnh_structure
-from pyaptamer.datasets._loaders._pfoa_loader import load_pfoa_structure
+from pyaptamer.datasets._loaders._pfoa import load_pfoa_structure
 
 __all__ = ["load_pfoa_structure", "load_1gnh_structure"]

pyaptamer/datasets/_loaders/_online_databank.py

@@ -3,7 +3,7 @@
 
 from Bio.PDB import PDBList
 
-from pyaptamer.utils.pdb_to_struct import pdb_to_struct
+from pyaptamer.utils import pdb_to_struct
 
 
 def load_from_rcsb(pdb_id, overwrite=False):

pyaptamer/utils/__init__.py

@@ -7,13 +7,15 @@
     "rna2vec",
     "pdb_to_struct",
     "struct_to_aaseq",
+    "pdb_to_aaseq",
 ]
 
+from pyaptamer.utils._pdb_to_aaseq import pdb_to_aaseq
+from pyaptamer.utils._pdb_to_struct import pdb_to_struct
 from pyaptamer.utils._rna import (
     dna2rna,
     encode_rna,
     generate_all_aptamer_triplets,
     rna2vec,
 )
-from pyaptamer.utils.pdb_to_struct import pdb_to_struct
-from pyaptamer.utils.struct_to_aaseq import struct_to_aaseq
+from pyaptamer.utils._struct_to_aaseq import struct_to_aaseq

pyaptamer/utils/_pdb_to_aaseq.py

@@ -0,0 +1,49 @@
+__author__ = "satvshr"
+__all__ = ["pdb_to_aaseq"]
+
+import os
+
+import pandas as pd
+from Bio import SeqIO
+
+
+def pdb_to_aaseq(pdb_file_path, return_type="list"):
+    """
+    Extract amino-acid sequences (SEQRES) from a PDB file.
+
+    Parameters
+    ----------
+    pdb_file_path : str or os.PathLike
+        Path to a PDB file.
+    return_type : {'list', 'pd.df'}, optional, default='list'
+        Format of the returned value:
+
+        - ``'list'`` : return a Python list of sequence strings (one per chain).
+        - ``'pd.df'`` : return a pandas.DataFrame indexed by chain id with a single
+        column ``'sequence'`` containing one-letter amino-acid strings.
+
+    Returns
+    -------
+    list of str or pandas.DataFrame
+        Depending on ``return_type``. If ``'list'``, returns a list of sequence
+        strings (one per SEQRES chain). If ``'pd.df'``, returns a DataFrame
+        where the index is the chain identifier when present (index name ``'chain'``)
+        and the column ``'sequence'`` contains the sequences. If no SEQRES records
+        are found, returns an empty list or empty DataFrame respectively.
+    """
+    pdb_path = os.fspath(pdb_file_path)
+    sequences = []
+    chains = []
+
+    with open(pdb_path) as handle:
+        for record in SeqIO.parse(handle, "pdb-seqres"):
+            sequences.append(str(record.seq))
+            chains.append(getattr(record, "id", None))
+
+    if return_type == "list":
+        return sequences
+
+    if return_type == "pd.df":
+        df = pd.DataFrame({"chain": chains, "sequence": sequences})
+        df = df.set_index("chain")
+        return df

pyaptamer/utils/tests/test_pdb_to_aaseq.py

@@ -0,0 +1,32 @@
+__author__ = "satvshr"
+
+import os
+
+from pyaptamer.utils import pdb_to_aaseq
+
+
+def test_pdb_to_aaseq():
+    """
+    Test that `pdb_to_aaseq` converts a PDB file path into a non-empty string
+    containing alphabetic characters.
+    """
+    pdb_file_path = os.path.join(
+        os.path.dirname(__file__), "..", "..", "datasets", "data", "1gnh.pdb"
+    )
+    sequences = pdb_to_aaseq(pdb_file_path)
+
+    assert isinstance(sequences, list), "pdb_to_aaseq should return a list"
+    assert len(sequences) > 0, "Returned list should not be empty"
+
+    for seq in sequences:
+        assert isinstance(seq, str), "Each entry should be a string"
+        assert len(seq) > 0, "Each sequence string should not be empty"
+
+    sequences = pdb_to_aaseq(pdb_file_path, return_type="pd.df")
+
+    assert not sequences.empty, "Returned DataFrame should not be empty"
+    assert "sequence" in sequences.columns, "DataFrame should have a 'sequence' column"
+
+    for seq in sequences["sequence"]:
+        assert isinstance(seq, str), "Each entry should be a string"
+        assert len(seq) > 0, "Each sequence string should not be empty"

pyaptamer/utils/tests/test_pdb_to_struct.py

@@ -2,7 +2,7 @@
 
 import os
 
-from pyaptamer.utils.pdb_to_struct import pdb_to_struct
+from pyaptamer.utils import pdb_to_struct
 
 
 def test_pdb_to_struct():

pyaptamer/utils/tests/test_struct_to_aaseq.py

@@ -1,7 +1,7 @@
 __author__ = "satvshr"
 
 from pyaptamer.datasets import load_1gnh_structure
-from pyaptamer.utils.struct_to_aaseq import struct_to_aaseq
+from pyaptamer.utils import struct_to_aaseq
 
 
 def test_struct_to_aaseq():