debase 0.6.0__py3-none-any.whl → 0.6.2__py3-none-any.whl

debase/lineage_format.py

@@ -30,6 +30,7 @@ from __future__ import annotations
 
 import argparse
 import csv
+import difflib
 import json
 import logging
 import os
@@ -726,7 +727,49 @@ Only include matches you are confident about. If no match exists, omit that enzy
                             gemini_matched_count += 1
                             log.info(f"Gemini matched '{substrate_id}' -> '{matched_id}' in campaign {campaign_id}")
                         else:
-                            log.warning(f"Gemini suggested match '{matched_id}' not found in sequence lookup")
+                            # Try fuzzy matching when exact match fails
+                            best_match = None
+                            best_score = 0
+                            
+                            # Try all possible keys in seq_lookup
+                            for key in seq_lookup.keys():
+                                if campaign_id in key:  # Only consider keys from same campaign
+                                    # Extract enzyme_id part from composite key
+                                    try:
+                                        _, key_enzyme_id = key.split('_', 1)
+                                    except ValueError:
+                                        continue
+                                    
+                                    # Calculate similarity score
+                                    score = difflib.SequenceMatcher(None, matched_id.lower(), key_enzyme_id.lower()).ratio()
+                                    
+                                    # Always track the highest score
+                                    if score > best_score:
+                                        best_score = score
+                                        best_match = key
+                            
+                            # Use the best match regardless of threshold (let user see the score)
+                            if best_match and best_score > 0.5:  # Lower threshold but log the score
+                                idx = entry["idx"]
+                                df.at[idx, "protein_sequence"] = seq_lookup[best_match]["aa_sequence"]
+                                df.at[idx, "aa_sequence"] = seq_lookup[best_match]["aa_sequence"]
+                                if seq_lookup[best_match]["nt_sequence"]:
+                                    df.at[idx, "nucleotide_sequence"] = seq_lookup[best_match]["nt_sequence"]
+                                    df.at[idx, "nt_sequence"] = seq_lookup[best_match]["nt_sequence"]
+                                
+                                # Also copy generation and parent_enzyme_id
+                                df.at[idx, "generation"] = seq_lookup[best_match]["generation"]
+                                df.at[idx, "parent_enzyme_id"] = seq_lookup[best_match]["parent_enzyme_id"]
+                                
+                                # Store the match for later mutation copying
+                                _, matched_enzyme = best_match.split('_', 1)
+                                df.at[idx, "_matched_enzyme_id"] = matched_enzyme
+                                df.at[idx, "_matched_campaign_id"] = campaign_id
+                                
+                                gemini_matched_count += 1
+                                log.info(f"Fuzzy matched '{substrate_id}' -> '{matched_enzyme}' (score: {best_score:.2f}) in campaign {campaign_id}")
+                            else:
+                                log.warning(f"No fuzzy match found for Gemini suggested '{matched_id}' in campaign {campaign_id} (best score: {best_score:.2f})")
                 
             except Exception as e:
                 log.warning(f"Failed to get Gemini matches for campaign {campaign_id}: {e}")