gsrap 0.9.0__py3-none-any.whl → 0.10.1__py3-none-any.whl

gsrap/parsedb/introduce.py

@@ -1,22 +1,27 @@
+import sys
+
 import pandas as pnd
 import cobra
 
-
 from .repeating import check_author
 from .repeating import check_rstring_arrow
 from .repeating import check_gpr
 from .repeating import add_reaction
+from .repeating import get_curator_notes
 
 from .manual import get_manual_sinks
 from .manual import get_manual_demands
 
 
     
-def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to_others, goodbefore, onlyauthor):
+def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to_others, outdir, goodbefore, onlyauthor):
     goodbefore_reached = False
+    
+    
+    logger.info("Parsing metabolites ('M' sheet)...")
 
     
-    logger.debug("Checking duplicated metabolite IDs...")
+    # check duplicated puremids:
     if len(set(db['M']['pure_mid'].to_list())) != len(db['M']): 
         pure_mids = db['M']['pure_mid'].to_list()
         duplicates = list(set([item for item in pure_mids if pure_mids.count(item) > 1]))
@@ -24,11 +29,12 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
         return 1
    
         
-    # parse M:
-    logger.debug("Parsing metabolites...")
+    # parse M (row by row):    
     db['M'] = db['M'].set_index('pure_mid', drop=True, verify_integrity=True)
     kc_ids_modeled = set()   # account for kc codes modeled
-    for pure_mid, row in db['M'].iterrows():
+    cnt = 0  # counter for parsed records
+    msg = '' # to be cleared
+    for iteration, (pure_mid, row) in enumerate(db['M'].iterrows()):
         
         
         # skip empty lines!
@@ -41,11 +47,11 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
         # manage goodbefore/onlyauthor
         if goodbefore != None and goodbefore_reached:
             if onlyauthor == None:
-                logger.info(f"Skipping metabolite '{pure_mid}' as requested with --goodbefore[0] '{goodbefore}'.")
+                logger.warning(f"Skipping metabolite '{pure_mid}' as requested with --goodbefore[0] '{goodbefore}'.")
                 continue
         
         
-        # parse author
+        # parse and get curators
         response = check_author(logger, pure_mid, row, db, 'M')
         if type(response) == int: return 1
         else: authors = response
@@ -55,7 +61,7 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
         if goodbefore != None and goodbefore_reached:
             if onlyauthor != None and onlyauthor not in authors:
                 authors_string = '; '.join(authors)
-                logger.info(f"Skipping metabolite '{pure_mid}' (authors '{authors_string}') as requested with --goodbefore[0] '{goodbefore}' and --onlyauthor '{onlyauthor}'.")
+                logger.warning(f"Skipping metabolite '{pure_mid}' (authors '{authors_string}') as requested with --goodbefore[0] '{goodbefore}' and --onlyauthor '{onlyauthor}'.")
                 continue
             
         
@@ -71,7 +77,7 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
             return 1
         
         
-        # check if 'kc' codes are real:
+        # parse kc:
         if pnd.isna(row['kc']): 
             logger.error(f"Metabolite '{pure_mid}' has missing KEGG annotation (kc): '{row['kc']}'.")
             return 1
@@ -79,11 +85,16 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
         kc_ids = [i.strip() for i in kc_ids]
         for kc_id in kc_ids:
             if kc_id == 'CXXXXX':  # not in KEGG; could be knowledge gap (e.g. methyl group acceptor in R10404)
-                logger.debug(f"Metabolite '{pure_mid}' is not in KEGG ('{kc_id}')!")
+                with open(f"{outdir}/logs/M.notkegg.txt", 'a') as f:
+                    print(f"Metabolite '{pure_mid}' is not in KEGG ('{kc_id}')!", file=f)
                 continue  
+            #    
+            # check if 'kc' codes are real:
             if kc_id not in idcollection_dict['kc']:
                 logger.error(f"Metabolite '{pure_mid}' has invalid KEGG annotation (kc): '{kc_id}'.")
                 return 1
+            #
+            # check if 'kc' was already used:
             if kc_id in kc_ids_modeled:
                 logger.error(f"KEGG annotation (kc) '{kc_id}' used in metabolite '{pure_mid}' is duplicated.")
                 return 1
@@ -95,6 +106,7 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
         if pnd.isna(row['inchikey']): 
             logger.error(f"Metabolite '{pure_mid}' has missing inchikey: '{row['inchikey']}'.")
             return 1
+        
         # check inchikey format:
         if len(row['inchikey']) != 27 or row['inchikey'][14] != '-' or row['inchikey'][25] != '-':
             logger.error(f"Metabolite '{pure_mid}' has badly formatted inchikey: '{row['inchikey']}'.")
@@ -109,7 +121,8 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
                     for eqbiggid in kegg_compound_to_others[kc_id]['bigg.metabolite']:
                         eqbiggids.add(eqbiggid)
         if pure_mid not in eqbiggids and eqbiggids != set():
-            logger.debug(f"Metabolites '{'; '.join(kc_ids)}' already in BiGG as {eqbiggids} ({authors} gave '{pure_mid}').")        
+            with open(f"{outdir}/logs/M.inbigg.txt", 'a') as f:
+                print(f"Metabolites '{'; '.join(kc_ids)}' already in BiGG as {eqbiggids} ({authors} gave '{pure_mid}').", file=f)        
         
         
         # add metabolite to model
@@ -122,6 +135,10 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
         m.compartment='c'
         
         
+        # write curators as annotations
+        m.annotation['curator_codes'] = authors
+        
+        
         # add annotations to model (same order of Memote)
         ankeys = [
             'pubchem.compound', 'kegg.compound', 'seed.compound',
@@ -132,53 +149,73 @@ def introduce_metabolites(logger, db, model, idcollection_dict, kegg_compound_to
         for ankey in ankeys:
             if ankey == 'kegg.compound': m.annotation[ankey] = set(kc_ids) - set(['CXXXXX'])
             else: m.annotation[ankey] = set()
+        #    
         # populate sets:
         for kc_id in kc_ids:
             if kc_id != 'CXXXXX':
                 if kc_id in kegg_compound_to_others.keys():
                     for ankey in ankeys:
                         m.annotation[ankey].update(kegg_compound_to_others[kc_id][ankey])
+        #
         # save as list: 
         for ankey in ankeys:
             m.annotation[ankey] = list(m.annotation[ankey])
             
             
-        # replace inchikey with manually-curated
+        # # force the manual-curated version of the inchikey
         if m.annotation['inchikey'] != [] and m.annotation['inchikey'] != [row['inchikey']]:
-            logger.debug(f"Metabolite '{pure_mid}': manual-curated inchikey ({[row['inchikey']]}) is diferent from the one derived from MNX ({m.annotation['inchikey']}).")
-        m.annotation['inchikey'] = [row['inchikey']]  # force the manual-curated version
+            with open(f"{outdir}/logs/M.diffinchi.txt", 'a') as f:
+                print(f"Metabolite '{pure_mid}': manual-curated inchikey ({[row['inchikey']]}) is different from the one derived from MNX ({m.annotation['inchikey']}).", file=f)
+        m.annotation['inchikey'] = [row['inchikey']]  
+        #
+        # remove inchikey if unknown:
         if m.annotation['inchikey'] == ['XXXXXXXXXXXXXX-XXXXXXXXXX-X']:
             m.annotation['inchikey'] = []
             
             
         # add SBO annotation
         m.annotation['sbo'] = ['SBO:0000247']  # generic metabolite
+        
+        
+        # add curator_notes
+        m.annotation['curator_notes'] = get_curator_notes(logger, row)
+        
+        
+        # communicate progress:
+        cnt += 1
+        msg = f"Done {cnt}/{len(db['M'])}!"
+        print(msg, file=sys.stderr, end='\r')
+    print(''.join([' ' for i in range(len(msg))]), file=sys.stderr, end='\r')
    
         
-    
+    # check goodbefore reaching:
     if goodbefore != None and goodbefore_reached == False:
-        logger.info(f"Metabolite '{goodbefore}' never reached. Are you sure about your --goodbefore?")
+        logger.warning(f"Metabolite '{goodbefore}' never reached. Are you sure about your --goodbefore?")
                     
                     
     return model
     
     
     
-def introduce_reactions(logger, db, model, idcollection_dict, kegg_reaction_to_others, goodbefore, onlyauthor): 
+def introduce_reactions(logger, db, model, idcollection_dict, kegg_reaction_to_others, outdir, goodbefore, onlyauthor): 
     goodbefore_reached = False
+    
+    
+    logger.info("Parsing non-transport reactions ('R' sheet)...")
                     
     
-    logger.debug("Checking duplicated reaction IDs...")
+    # check duplicated rids:
     if len(set(db['R']['rid'].to_list())) != len(db['R']): 
-        pure_mids = db['R']['rid'].to_list()
-        duplicates = list(set([item for item in pure_mids if pure_mids.count(item) > 1]))
+        rids = db['R']['rid'].to_list()
+        duplicates = list(set([item for item in rids if rids.count(item) > 1]))
         logger.error(f"Sheet 'R' has duplicated reactions: {duplicates}.")
         return 1
     
         
-    # parse R:
-    logger.debug("Parsing reactions...")
+    # parse R (row by row):
     db['R'] = db['R'].set_index('rid', drop=True, verify_integrity=True)
+    cnt = 0  # counter for parsed records
+    msg = '' # to be cleared
     for rid, row in db['R'].iterrows():
         
         
@@ -192,11 +229,11 @@ def introduce_reactions(logger, db, model, idcollection_dict, kegg_reaction_to_o
         # manage goodbefore/onlyauthor
         if goodbefore != None and goodbefore_reached:
             if onlyauthor == None:
-                logger.info(f"Skipping reaction '{rid}' as requested with --goodbefore[1] '{goodbefore}'.")
+                logger.warning(f"Skipping reaction '{rid}' as requested with --goodbefore[1] '{goodbefore}'.")
                 continue
         
         
-        # parse author
+        # parse and get curators
         response = check_author(logger, rid, row, db, 'R')
         if type(response) == int: return 1
         else: authors = response
@@ -206,7 +243,7 @@ def introduce_reactions(logger, db, model, idcollection_dict, kegg_reaction_to_o
         if goodbefore != None and goodbefore_reached:
             if onlyauthor != None and onlyauthor not in authors:
                 authors_string = '; '.join(authors)
-                logger.info(f"Skipping reaction '{rid}' (authors '{authors_string}') as requested with --goodbefore[1] '{goodbefore}' and --onlyauthor '{onlyauthor}'.")
+                logger.warning(f"Skipping reaction '{rid}' (authors '{authors_string}') as requested with --goodbefore[1] '{goodbefore}' and --onlyauthor '{onlyauthor}'.")
                 continue
         
         
@@ -215,7 +252,7 @@ def introduce_reactions(logger, db, model, idcollection_dict, kegg_reaction_to_o
         if response == 1: return 1
         
 
-        # check if 'kr' codes are real:
+        # parse 'kr':
         if pnd.isna(row['kr']): 
             logger.error(f"Reaction '{rid}' has missing KEGG annotation (kr): '{row['kr']}'.")
             return 1
@@ -223,15 +260,18 @@ def introduce_reactions(logger, db, model, idcollection_dict, kegg_reaction_to_o
         kr_ids = [i.strip() for i in kr_ids]
         for kr_id in kr_ids:
             if kr_id == 'RXXXXX':  # not in KEGG; could be knowledge gap 
-                logger.debug(f"Reaction '{rid}' is not in KEGG ('{kr_id}')!")
+                with open(f"{outdir}/logs/R.notkegg.txt", 'a') as f:
+                    print(f"Reaction '{rid}' is not in KEGG ('{kr_id}')!", file=f)
                 continue  
+            #
+            # check if 'kr' codes are real:
             if kr_id not in idcollection_dict['kr']:
                 logger.error(f"Reaction '{rid}' has invalid KEGG annotation (kr): '{kr_id}'.")
                 return 1
         
             
         # check GPR:
-        response = check_gpr(logger, rid, row, kr_ids, idcollection_dict, 'R')
+        response = check_gpr(logger, rid, row, kr_ids, idcollection_dict, 'R', outdir)
         if response == 1: return 1
     
     
@@ -243,28 +283,40 @@ def introduce_reactions(logger, db, model, idcollection_dict, kegg_reaction_to_o
                     for eqbiggid in kegg_reaction_to_others[kr_id]['bigg.reaction']:
                         eqbiggids.add(eqbiggid)
         if rid not in eqbiggids and eqbiggids != set():
-            logger.debug(f"Reactions '{'; '.join(kr_ids)}' already in BiGG as {eqbiggids} ({authors} gave '{rid}').") 
+            with open(f"{outdir}/logs/R.inbigg.txt", 'a') as f:
+                print(f"Reactions '{'; '.join(kr_ids)}' already in BiGG as {eqbiggids} ({authors} gave '{rid}').", file=f) 
         
         
         # add reaction to model
-        response = add_reaction(logger, model, rid, row, kr_ids, kegg_reaction_to_others, 'R')
+        response = add_reaction(logger, model, rid, authors, row, kr_ids, kegg_reaction_to_others, 'R', outdir)
         if response == 1: return 1
+    
+    
+        # communicate progress:
+        cnt += 1
+        msg = f"Done {cnt}/{len(db['R'])}!"
+        print(msg, file=sys.stderr, end='\r')
+    print(''.join([' ' for i in range(len(msg))]), file=sys.stderr, end='\r')
                
     
+    # check goodbefore reaching:
     if goodbefore != None and goodbefore_reached == False:
-        logger.info(f"Reaction '{goodbefore}' never reached. Are you sure about your --goodbefore?")
+        logger.warning(f"Reaction '{goodbefore}' never reached. Are you sure about your --goodbefore?")
     
     
     return model
       
     
     
-def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_to_others, goodbefore, onlyauthor): 
+def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_to_others, outdir, goodbefore, onlyauthor): 
     goodbefore_reached = False
-                    
+    
+    
+    logger.info("Parsing transport reactions ('T' sheet)...")
     
     
     def clone_to_external(model, mid_c, mid_e):
+        # given an existing '_c' M, create its '_e' equivalent
     
         m = cobra.Metabolite(f'{mid_e}')
         model.add_metabolites([m])
@@ -281,10 +333,12 @@ def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_t
             
     
     def add_exchange_reaction(model, mid_e):
+        # given an existing '_e' M, create the corresponding EX_change reaction
         
         r = cobra.Reaction(f'EX_{mid_e}')
         model.add_reactions([r])
         r = model.reactions.get_by_id(f'EX_{mid_e}')
+        
         r.name = f"Exchange for {model.metabolites.get_by_id(mid_e).name}"
         r.build_reaction_from_string(f'{mid_e} --> ')
         r.bounds = (0, 1000)
@@ -294,22 +348,15 @@ def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_t
         
     
     
-    
-    # get all already inserted metabolites
+    # get all already inserted metabolites and reactions
     mids_parsed = [m.id for m in model.metabolites]
     rids_parsed = [r.id for r in model.reactions]
     
     
-    # protons may not have an explicit transporter
-    clone_to_external(model, 'h_c', 'h_e')
-    mids_parsed.append('h_e')
-    add_exchange_reaction(model, 'h_e')
-    rids_parsed.append(f'EX_h_e')
-    
-    
-    # parse T:
-    logger.debug("Parsing transporters...")
+    # parse T (row by row):
     db['T'] = db['T'].set_index('rid', drop=True, verify_integrity=True)
+    cnt = 0  # counter for parsed records
+    msg = '' # to be cleared
     for rid, row in db['T'].iterrows():
         
         
@@ -320,10 +367,16 @@ def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_t
             goodbefore_reached = True
             
             
+        # avoid duplicates!
+        if rid in rids_parsed:
+            logger.error(f"Tranport '{rid}' has ID identical to previously added reaction!")
+            return 1
+            
+            
         # manage goodbefore/onlyauthor
         if goodbefore != None and goodbefore_reached:
             if onlyauthor == None:
-                logger.info(f"Skipping transport '{rid}' as requested with --goodbefore[2] '{goodbefore}'.")
+                logger.warning(f"Skipping transport '{rid}' as requested with --goodbefore[2] '{goodbefore}'.")
                 continue
         
         
@@ -337,7 +390,7 @@ def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_t
         if goodbefore != None and goodbefore_reached:
             if onlyauthor != None and onlyauthor not in authors:
                 authors_string = '; '.join(authors)
-                logger.info(f"Skipping transport '{rid}' (authors '{authors_string}') as requested with --goodbefore[2] '{goodbefore}' and --onlyauthor '{onlyauthor}'.")
+                logger.warning(f"Skipping transport '{rid}' (authors '{authors_string}') as requested with --goodbefore[2] '{goodbefore}' and --onlyauthor '{onlyauthor}'.")
                 continue
         
         
@@ -346,7 +399,7 @@ def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_t
         if response == 1: return 1
     
     
-        # check if 'kr' codes are real:
+        # parse 'kr':
         if pnd.isna(row['kr']): 
             logger.error(f"Reaction '{rid}' has missing KEGG annotation (kr): '{row['kr']}'.")
             return 1
@@ -354,37 +407,39 @@ def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_t
             kr_ids = row['kr'].split(';')
             kr_ids = [i.strip() for i in kr_ids]
             for kr_id in kr_ids:
+                #
+                # check if 'kr' codes are real
                 if kr_id not in idcollection_dict['kr']:
                     logger.error(f"Reaction '{rid}' has invalid KEGG annotation (kr): '{kr_id}'.")
                     return 1
-        else: kr_ids = []
+        else: 
+            # no 'kr' for the majority of transport reactions!
+            kr_ids = []
 
             
         # check GPR:
-        response = check_gpr(logger, rid, row, kr_ids, idcollection_dict, 'T')
+        response = check_gpr(logger, rid, row, kr_ids, idcollection_dict, 'T', outdir)
         if response == 1: return 1
         
         
-        # get involved metabolites:
-        involved_mids = row['rstring'].split(' ')
-        involved_mids = [i for i in involved_mids if i not in ['-->', '<=>']]
-        
-        
-        # the external metabolite must be already modeled as cytosolic
+        # iterate the involved metabolites
+        involved_mids = row['rstring'].split(' ')  # dirty (arrows, coefficints are included)
         for mid in involved_mids: 
             if mid.endswith('_e'):
                 mid_e = mid
                 mid_c = mid.rsplit('_', 1)[0] + '_c'
+                #
+                # the cytosolic counterpart must be already modeled:
                 if mid_c not in mids_parsed:
                     logger.error(f"{rid}: the metabolite '{mid_c}', counterpart of '{mid_e}', was not previously modeled.")
                     return 1
-                
-                # add external metabolite to model
+                #
+                # clone to add external metabolite to model
                 if mid_e not in mids_parsed:
                     clone_to_external(model, mid_c, mid_e)
                     mids_parsed.append(mid_e)
-                    
-                # add exchange reaction to model
+                #    
+                # add corresponding exchange reaction to model
                 if f'EX_{mid_e}' not in rids_parsed:
                     add_exchange_reaction(model, mid_e)
                     rids_parsed.append(f'EX_{mid_e}')
@@ -398,16 +453,26 @@ def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_t
                     for eqbiggid in kegg_reaction_to_others[kr_id]['bigg.reaction']:
                         eqbiggids.add(eqbiggid)
         if rid not in eqbiggids and eqbiggids != set():
-            logger.debug(f"Reactions '{'; '.join(kr_ids)}' already in BiGG as {eqbiggids} ({authors} gave '{rid}').") 
+            with open(f"{outdir}/logs/T.inbigg.txt", 'a') as f:
+                print(f"Reactions '{'; '.join(kr_ids)}' already in BiGG as {eqbiggids} ({authors} gave '{rid}').", file=f) 
                     
                     
         # add reaction to model
-        response = add_reaction(logger, model, rid, row, kr_ids, kegg_reaction_to_others, 'T')
+        response = add_reaction(logger, model, rid, authors, row, kr_ids, kegg_reaction_to_others, 'T', outdir)
         if response == 1: return 1
+        rids_parsed.append(rid)   # update list of rids in model
+        
         
+        # communicate progress:
+        cnt += 1
+        msg = f"Done {cnt}/{len(db['T'])}!"
+        print(msg, file=sys.stderr, end='\r')
+    print(''.join([' ' for i in range(len(msg))]), file=sys.stderr, end='\r')
         
+        
+    # check goodbefore reaching:
     if goodbefore != None and goodbefore_reached == False:
-        logger.info(f"Transport '{goodbefore}' never reached. Are you sure about your --goodbefore?")
+        logger.warning(f"Transport '{goodbefore}' never reached. Are you sure about your --goodbefore?")
         
     
     return model
@@ -416,6 +481,10 @@ def introduce_transporters(logger, db, model, idcollection_dict, kegg_reaction_t
 
 def introduce_sinks_demands(logger, model): 
     
+    
+    logger.debug("Introducing sinks and demands...")
+    
+    
     sinks = get_manual_sinks()
     demands = get_manual_demands()
     

gsrap/parsedb/manual.py

@@ -15,8 +15,27 @@ def get_krs_to_exclude():
     return set([
         'R12328', 'R05190',  # general forms of fatty acid biosynthesis
         'R01347', 'R01348', 'R04121',  # general forms of fatty acid degradation
+        'R11671',  # multi-step fatty acids reactions
+        'R07860', 'R01317', 'R07064',  # aspecific fatty acid reactions
+        'R11311', 'R11256', 'R11308', 'R08772', 'R08770',  # polymer reactions
+        
+        # inconclusive due to semplification
+        'R12425', 
+        
+        # "incomplete reaction" / "unclear reaction"
+        'R08414', 'R13037', 'R13034', 'R13036', 'R02825', 'R11178', 'R13325', 'R12855', 'R12856', 'R09809', 
+        'R09808', 'R08035', 'R08034', 'R11470', 'R09360', 'R08139', 'R08318', 'R07859', 'R09361', 'R09349', 
+        'R13149', 'R13066', 'R11467', 'R11255', 'R08986', 'R13156', 'R13074', 'R13150', 'R11302', 'R11388', 
+        'R08341', 'R13147', 'R13155', 'R08339', 'R11466', 'R08272', 'R09348', 'R09362', 'R11107', 'R08340', 
+        'R07940', 'R11120', 'R11245', 'R08269', 'R11131', 'R07943', 'R08342', 'R06766', 'R12584', 'R09852',
+        'R08268', 'R11129', 'R06702', 'R08866', 'R12555', 'R08927', 'R08343', 'R13067', 'R13069', 'R13068',
+        'R05670', 'R06694', 'R09851', 'R11465', 'R08928', 'R11389', 'R11464', 'R13087', 'R12586', 'R11304', 
+        'R08984', 'R11254', 'R13165', 'R12884', 'R08865', 'R13151', 'R08132', 'R08929', 'R06701', 'R08345',
+        'R11365', 'R11303', 'R06670', 'R11364', 'R09347', 'R08293', 'R11362', 'R03872', 'R06339', 'R10481', 
+        'R10480', 'R13341', 'R06505', 'R06504', 'R06326', 'R06470', 'R06467', 'R06327', 'R06503', 'R09847',
+        'R13479', 'R13447', 'R13478', 'R07510', 'R04546', 'R06468', 'R05624', 'R10706', 'R13454', 'R13556',
+        'R13455', 'R12691', 
     ])
-    
 
 
 
@@ -29,12 +48,13 @@ def get_rids_with_mancheck_gpr():
     return rids_mancheck_gpr
 
 
+
 def get_rids_with_mancheck_balancing():
     rids_mancheck_bal = [  # same reactions involving ATP can be reversible
         
         # SECTION "reversible both in KEGG and MetaCyc"
         'PGK', 'SUCOAS', 'ADK1', 'GK1', 'NNATr', 'CYTK1', 'ACKr',
-        'DGK1', 'PPAKr', 'ATPSr', 'NDPK10',
+        'DGK1', 'PPAKr', 'ATPSr', 'NDPK10', 'BUTKr',
         
         ### SECTION "reversible in KEGG but not in MetaCyc" ###
         'CYTK2',  # clearly reversible in KEGG but not in MetaCyc (RXN-7913)