debase 0.5.0__tar.gz → 0.6.0__tar.gz

{debase-0.5.0/src/debase.egg-info → debase-0.6.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: debase
-Version: 0.5.0
+Version: 0.6.0
 Summary: Enzyme lineage analysis and sequence extraction package
 Home-page: https://github.com/YuemingLong/DEBase
 Author: DEBase Team

{debase-0.5.0 → debase-0.6.0}/src/debase/_version.py

@@ -1,3 +1,3 @@
 """Version information."""
 
-__version__ = "0.5.0"
+__version__ = "0.6.0"

{debase-0.5.0 → debase-0.6.0}/src/debase/lineage_format.py

@@ -35,7 +35,6 @@ import logging
 import os
 import pickle
 import re
-import sqlite3
 import sys
 import time
 from concurrent.futures import ThreadPoolExecutor, as_completed
@@ -137,8 +136,7 @@ SUBSTRATE_CACHE_FILE: Path = CACHE_DIR / "substrate_smiles_cache.pkl"
 CANONICAL_CACHE_FILE: Path = CACHE_DIR / "canonical_smiles_cache.pkl"
 CACHE_DIR.mkdir(parents=True, exist_ok=True)
 
-# Local PubChem DB (optional) --------------------------------------------------------
-PUBCHEM_DB_PATH: Path = Path(__file__).parent.parent.parent / "data" / "iupac2smiles.db"
+# API endpoints for IUPAC to SMILES conversion --------------------------------------
 
 # Gemini API configuration -----------------------------------------------------------
 GEMINI_API_KEY: str = os.environ.get("GEMINI_API_KEY", "")
@@ -323,37 +321,7 @@ SUBSTRATE_CACHE: Dict[str, str] = _load_pickle(SUBSTRATE_CACHE_FILE)
 CANONICAL_CACHE: Dict[str, str] = _load_pickle(CANONICAL_CACHE_FILE)
 
 
-# --- Database lookup ---------------------------------------------------------------
-class PubChemDB:
-    """Very thin wrapper around a local SQLite mapping IUPAC -> SMILES."""
-
-    def __init__(self, path: Path | str) -> None:
-        self.path = Path(path)
-        self._conn: Optional[sqlite3.Connection] = None
-        if not self.path.exists():
-            log.warning("Local PubChem DB not found at %s", self.path)
-
-    def _connect(self) -> sqlite3.Connection:
-        if self._conn is None:
-            self._conn = sqlite3.connect(str(self.path))
-        return self._conn
-
-    def lookup(self, name: str) -> Optional[str]:
-        if not self.path.exists():
-            return None
-        sql = "SELECT smiles FROM x WHERE name = ? LIMIT 1"
-        try:
-            # Create a new connection for thread safety
-            conn = sqlite3.connect(str(self.path))
-            cur = conn.execute(sql, (name.lower(),))
-            row = cur.fetchone()
-            conn.close()
-            return row[0] if row else None
-        except Exception:  # pragma: no cover
-            return None
-
-
-PC_DB = PubChemDB(PUBCHEM_DB_PATH)
+# --- Removed local database - using only online APIs -------------------------------
 
 
 # === 5. SEQUENCE / MUTATION HELPERS ================================================
@@ -481,12 +449,7 @@ def _name_to_smiles(name: str, is_substrate: bool) -> str:
     if not name or name.lower() in ['nan', 'none', 'null', 'n/a', 'na', '']:
         return ""
     
-    # 1. Local DB (fast, offline)
-    db_smiles = PC_DB.lookup(name)
-    if db_smiles:
-        return db_smiles
-
-    # 2. OPSIN (if installed) ---------------------------------------------------
+    # 1. OPSIN (if installed) - fast and reliable for IUPAC names
     try:
         import subprocess
 
@@ -503,12 +466,7 @@ def _name_to_smiles(name: str, is_substrate: bool) -> str:
     except FileNotFoundError:
         pass  # OPSIN not installed
 
-    # 3. Gemini search (for complex compounds) ---------------------------------
-    gemini_smiles = search_smiles_with_gemini(name)
-    if gemini_smiles:
-        return gemini_smiles
-    
-    # 4. PubChem PUG REST (online) ---------------------------------------------
+    # 2. PubChem PUG REST API (online) - comprehensive database
     try:
         import requests
 
@@ -521,6 +479,11 @@ def _name_to_smiles(name: str, is_substrate: bool) -> str:
             return pug_smiles
     except Exception:  # pragma: no cover
         pass
+    
+    # 3. Gemini search (for complex compounds) - AI fallback
+    gemini_smiles = search_smiles_with_gemini(name)
+    if gemini_smiles:
+        return gemini_smiles
 
     # Return empty string if all methods fail
     return ""

{debase-0.5.0 → debase-0.6.0}/src/debase/reaction_info_extractor.py

@@ -1195,7 +1195,8 @@ class ReactionExtractor:
                 # Create a flexible pattern that handles various spacing and formatting
                 # This pattern looks for "Figure" (case insensitive) followed by optional spaces
                 # then the figure number, then any of: period, colon, space+capital letter, or end of line
-                flexible_pattern = rf"(?i)figure\s*{re.escape(main_figure_num)}(?:\.|:|(?=\s+[A-Z])|\s*$)"
+                # Also match at the beginning of a line to catch captions
+                flexible_pattern = rf"(?i)(?:^|\n)\s*figure\s*{re.escape(main_figure_num)}(?:\.|:|(?=\s+[A-Z])|\s*$)"
                 
                 LOGGER.debug("Looking for figure caption '%s' with flexible pattern: %s", 
                            main_figure_num, flexible_pattern)
@@ -1231,11 +1232,17 @@ class ReactionExtractor:
                         continue
                         
                     # Check if the remaining text looks like a caption (contains descriptive words)
+                    # Expanded list of caption keywords to be more inclusive
                     first_words = remaining_text[:50].lower()
-                    if not any(word in first_words for word in ['detailed', 'representative', 'shows', 'comparison', 
-                                                                 'illustrates', 'demonstrates', 'results', 'data',
-                                                                 'chromatogram', 'spectra', 'analysis', 'site-directed',
-                                                                 'mutagenesis', 'mutants']):
+                    caption_keywords = ['detailed', 'representative', 'shows', 'comparison', 
+                                      'illustrates', 'demonstrates', 'results', 'data',
+                                      'chromatogram', 'spectra', 'analysis', 'site-directed',
+                                      'mutagenesis', 'mutants', 'evolution', 'directed',
+                                      'screening', 'reaction', 'variant', 'enzyme', 'protein',
+                                      'activity', 'performance', 'yield', 'selectivity',
+                                      'characterization', 'optimization', 'development',
+                                      'structure', 'domain', 'crystal', 'model']
+                    if not any(word in first_words for word in caption_keywords):
                         LOGGER.debug("Skipping: doesn't look like caption text: %s", first_words)
                         continue
                     
@@ -1322,6 +1329,39 @@ class ReactionExtractor:
                     self._figure_cache.put(cache_key, result)
                     return result
         
+        # Fallback: If no caption found, try to find any page that mentions this figure
+        LOGGER.info("No figure caption found for '%s', trying fallback search", ref)
+        
+        for doc_idx, doc in enumerate(docs):
+            doc_name = "MS" if doc_idx == 0 else "SI"
+            for page_number in range(doc.page_count):
+                page = doc.load_page(page_number)
+                page_text = page.get_text()
+                
+                # Look for any mention of the figure reference
+                if re.search(rf'\b{re.escape(ref)}\b', page_text, re.IGNORECASE):
+                    LOGGER.info("Found '%s' mentioned on page %d of %s document (fallback)", 
+                               ref, page_number + 1, doc_name)
+                    
+                    # Extract the entire page as the figure might be on this page
+                    mat = fitz.Matrix(5.0, 5.0)  # 5x zoom for better quality
+                    pix = page.get_pixmap(matrix=mat)
+                    pix = self._ensure_rgb_pixmap(pix)
+                    img_bytes = pix.tobytes("png")
+                    
+                    # Save PNG to debug directory if available
+                    if self.debug_dir:
+                        timestamp = int(time.time())
+                        png_file = self.debug_dir / f"fallback_{ref.replace(' ', '_')}_{timestamp}.png"
+                        with open(png_file, 'wb') as f:
+                            f.write(img_bytes)
+                        LOGGER.info("Saved fallback page image to: %s", png_file)
+                    
+                    result = b64encode(img_bytes).decode()
+                    # Cache the result
+                    self._figure_cache.put(cache_key, result)
+                    return result
+        
         LOGGER.warning("_extract_page_png returning None for '%s' - figure not found in any document", ref)
         return None
     

{debase-0.5.0 → debase-0.6.0/src/debase.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: debase
-Version: 0.5.0
+Version: 0.6.0
 Summary: Enzyme lineage analysis and sequence extraction package
 Home-page: https://github.com/YuemingLong/DEBase
 Author: DEBase Team