Refactor fetch_refseq_sequence to improve error handling and input validation

- Updated the fetch_refseq_sequence function to handle HTTP errors more robustly by distinguishing between urllib and requests exceptions.
- Added input validation to skip empty or NaN RefSeq IDs.
- Cleaned up whitespace and comma-separated values from the RefSeq ID.
- Enhanced error messages for better debugging.
- Updated exception handling in _fetch_sequence_for_row to accommodate new error handling in fetch_refseq_sequence.

notebooks/01_EDA_Psort.ipynb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f99370bb677795f54b6778596ada73015381847b1838e3cc22553ede7a03dbc3
-size 10363061
+oid sha256:6c455bf2fd2e35c927e325cae70eb912d288e69801397c163615a7931651a3ce
+size 15152718

src/my_utils.py

@@ -32,7 +32,8 @@ from sklearn.base import BaseEstimator
 import umap
 
 import requests
-from requests.exceptions import HTTPError
+from urllib.error import HTTPError as URLLibHTTPError
+from requests.exceptions import HTTPError as RequestsHTTPError
 from Bio import Entrez
 from Bio import SeqIO
 from tqdm import tqdm
@@ -524,20 +525,30 @@ def fetch_uniprot_sequence(uniprot_id: str):
     else:
         print(f'URL inválido o no accesible: {url}')
 
-def fetch_refseq_sequence(refseq_id : str):
+from Bio import Entrez, SeqIO
+
+from io import StringIO
+import pandas as pd
+import requests
+
+def fetch_refseq_sequence(refseq_id: str) -> str | None:
     """
     Fetch the protein sequence for the given RefSeq ID using NCBI Entrez.
-    Returns the raw amino-acid sequence as a string.
+    Returns the raw amino-acid sequence as a string, or None on failure.
     """
-
-    Entrez.email = "puglia.jd@gmail.com"   # REQUIRED
+    # ——— NCBI credentials ———
+    Entrez.email   = "puglia.jd@gmail.com"
     Entrez.api_key = "d768134734612d58be85117e1ff22e243807"
-    # Check if the ID is NaN or None
-    if pd.isna(refseq_id) or refseq_id is None:
+
+    # ——— Validate input ———
+    if not refseq_id or pd.isna(refseq_id):
+        print(f"[SKIP] Empty or NaN RefSeq ID: `{refseq_id}`")
         return None
-    
-    fasta_data = None
-    
+
+    # Clean up possible whitespace or comma‐separated values
+    refseq_id = str(refseq_id).strip().split(",")[0]
+
+    # ——— 1) Try NCBI Entrez ———
     try:
         handle = Entrez.efetch(
             db="protein",
@@ -548,18 +559,26 @@ def fetch_refseq_sequence(refseq_id : str):
         record = SeqIO.read(handle, "fasta")
         handle.close()
         return str(record.seq)
-    except (HTTPError, ValueError):
 
+    except (URLLibHTTPError, ValueError) as e:
+        # catches bad HTTP status and parsing errors
+        print(f"[Entrez] Failed for `{refseq_id}`: {e}")
+
+    # ——— 2) Fallback: RCSB FASTA page ———
+    try:
         url = f"https://www.rcsb.org/fasta/entry/{refseq_id}"
         response = requests.get(url, timeout=60)
-        if response.status_code == 200:
-            try:
-                fasta_data = response.text
-                fasta_io = StringIO(fasta_data)
-                record = SeqIO.read(fasta_io, "fasta")
-                return str(record.seq)
-            except ValueError:
-                print(f"No se pudo convertir {fasta_data}, id: {refseq_id}")
+        response.raise_for_status()  # raises RequestsHTTPError on 4xx/5xx
+        fasta_io = StringIO(response.text)
+        record = SeqIO.read(fasta_io, "fasta")
+        return str(record.seq)
+
+    except (RequestsHTTPError, ValueError) as e2:
+        print(f"[RCSB] Failed for `{refseq_id}`: {e2}")
+
+    # ——— All methods failed ———
+    return None
+
 
 
 def _fetch_sequence_for_row(idx, row):
@@ -572,7 +591,7 @@ def _fetch_sequence_for_row(idx, row):
     if swiss_id and not pd.isna(swiss_id):
         try:
             sequence = fetch_uniprot_sequence(swiss_id)
-        except HTTPError as e:
+        except (URLLibHTTPError, RequestsHTTPError) as e:
             print(f"Warning: SwissProt fetch failed for {swiss_id} with HTTP {e}")
             sequence = None
 
@@ -580,7 +599,7 @@ def _fetch_sequence_for_row(idx, row):
     if not sequence and row.get('Refseq_Accession') and not pd.isna(row['Refseq_Accession']):
         try:
             sequence = fetch_refseq_sequence(row['Refseq_Accession'])
-        except HTTPError as e:
+        except (URLLibHTTPError, RequestsHTTPError) as e:
             print(f"Warning: RefSeq fetch failed for {row['Refseq_Accession']} with HTTP {e}")
             sequence = None
 
@@ -588,7 +607,7 @@ def _fetch_sequence_for_row(idx, row):
     if not sequence and row.get('Other_Accession') and not pd.isna(row['Other_Accession']):
         try:
             sequence = fetch_refseq_sequence(row['Other_Accession'])
-        except HTTPError as e:
+        except (URLLibHTTPError, RequestsHTTPError) as e:
             print(f"Warning: RefSeq fetch failed for {row['Other_Accession']} with HTTP {e}")
             sequence = None