Merge pull request #731 from JHL-452b/feat_smiles_inchi

Feat: added smiles and inchi of metabolites

code/modelCuration/getSmiles.py

@@ -0,0 +1,197 @@
+import pandas as pd
+df = pd.read_excel('./Human-GEM.xlsx',sheet_name='METS')
+df1 = pd.read_excel('./metabolites_smiles_.xlsx')
+df2 = pd.merge(df,df1,on='REPLACEMENT ID',how='left')
+df2.to_excel('./metabolites_smiles.xlsx',index=False)
+
+
+import pandas as pd
+df = pd.read_excel('./metabolites_smiles.xlsx')
+df1 = pd.read_csv('./database/mnx_chem_depr.tsv', sep='\t')
+id = df1['deprecated_ID'].tolist() 
+df['metMetaNetXID_new'] = ''
+for i in range(len(df)):
+    if df['metMetaNetXID'][i] in id:
+        index = id.index(df['metMetaNetXID'][i])
+        df['metMetaNetXID_new'][i] = df1['ID'][index]
+df.to_excel('./metabolites_smiles.xlsx', index=False)
+
+
+#get SMILES from database and model
+import pandas as pd
+from tqdm import tqdm
+#df = pd.read_csv('./database/chebi_id_smiles.csv')
+df = pd.read_csv('./database/kegg_compound.txt', sep='\t')
+#df = pd.read_csv('./database/chebi_second_id_smiles.csv')
+#df = pd.read_csv('./database/recon3d_smiles.csv')
+
+#id = df['ChEBI ID'].to_list()
+id = df['KEGG'].to_list()
+#id = df['Secondary ChEBI ID'].to_list()
+#id = df['metRecon3DID'].to_list()
+df1 = pd.read_excel('./metabolites_smiles.xlsx')
+#s = 'metChEBIID'
+s = 'metKEGGID'
+#s = 'metRecon3DID'
+#df1['SMILES'] = None
+for i in tqdm(range(len(df1)),total=len(df1)):
+    if df1['SMILES'].isna()[i] == False:
+        continue
+    try:
+        if i == 0:
+            if df1[s].isna()[i] == True:
+                continue
+            else:
+                if df1[s][i] in id:
+                    index = id.index(df1[s][i])
+                    df1['SMILES'][i] = df['SMILES'][index]
+
+        else:
+            if df1[s][i] == df1[s][i-1]:
+                continue
+            else:
+                if df1[s].isna()[i] == True:
+                    continue
+                else:
+                    if df1[s][i] in id:
+                        index = id.index(df1[s][i])
+                        df1['SMILES'][i] = df['SMILES'][index]
+    except:
+        continue
+
+for i in tqdm(range(len(df1)),total=len(df1)):
+    if i != 0:
+        if df1[s][i] == df1[s][i-1]:
+            df1['SMILES'][i] = df1['SMILES'][i-1]
+
+df1.to_excel('metabolites_smiles.xlsx', index=False)
+
+
+#get SMILES from database and model
+import pandas as pd
+from tqdm import tqdm
+
+df = pd.read_csv('./database/mnx_chem_prop.tsv', sep='\t')
+id = df['id'].to_list()
+
+df1 = pd.read_excel('./metabolites_smiles.xlsx')
+s = 'metMetaNetXID'
+for i in tqdm(range(len(df1)),total=len(df1)):
+    if df1['SMILES'].isna()[i] == False:
+        continue
+    try:
+        if i == 0:
+            if df1[s].isna()[i] == True:
+                continue
+            else:
+                for j in df1[s][i].split(';'):
+                    if j in id:
+                        index = id.index(j)
+                        df1['SMILES'][i] = df['SMILES'][index]
+                        break
+
+        else:
+            if df1[s][i] == df1[s][i-1]:
+                continue
+            else:
+                if df1[s].isna()[i] == True:
+                    continue
+                else:
+
+                    for j in df1[s][i].split(';'):
+                        if j in id:
+                            index = id.index(j)
+                            df1['SMILES'][i] = df['SMILES'][index]
+                            break
+    except:
+        continue
+
+for i in tqdm(range(len(df1)),total=len(df1)):
+    if i != 0:
+        if df1[s][i] == df1[s][i-1]:
+            df1['SMILES'][i] = df1['SMILES'][i-1]
+
+df1.to_excel('metabolites_smiles.xlsx', index=False)
+
+
+import pubchempy as pcp
+
+import pandas as pd
+from tqdm import tqdm
+failed = []
+df1 = pd.read_excel('./metabolites_smiles.xlsx')
+#df1['SMILES'] = ''
+for i in tqdm(range(len(df1)),total=len(df1)):
+    if df1['SMILES'].isna()[i] == False:
+        continue
+    try:
+        #df1['metChEBIID'][i] = df1['metChEBIID'][i].replace('CHEBI:','')
+        if i == 0:
+            if df1['metPubChemID'].isna()[i] == True:
+                continue
+            else:
+                cid = int(df1['metPubChemID'][i])
+                df1['SMILES'][i] = pcp.Compound.from_cid(cid).isomeric_smiles
+        else:
+            if df1['metsNoComp'][i] == df1['metsNoComp'][i-1]:
+                continue
+            else:
+                if df1['metPubChemID'].isna()[i] == True:
+                    continue
+                else:
+                    cid = int(df1['metPubChemID'][i])
+                    df1['SMILES'][i] = pcp.Compound.from_cid(cid).isomeric_smiles
+                    #print(df1['SMILES'][i])
+    except:
+        failed.append(df1['metPubChemID'][i])
+        continue
+for i in tqdm(range(len(df1)),total=len(df1)):
+    if i != 0:
+        if df1[s][i] == df1[s][i-1]:
+            df1['SMILES'][i] = df1['SMILES'][i-1]
+df1.to_excel('./metabolites_smiles.xlsx', index=False)
+# Get the SMILES string of the compound
+
+
+from rdkit import Chem
+def standardize_smiles(smiles):
+    mol = Chem.MolFromSmiles(smiles)
+    if mol is None:
+        return None
+    return Chem.MolToSmiles(mol)
+df = pd.read_excel('metabolites_smiles.xlsx')
+df['standard_smiles'] = df['SMILES'].apply(standardize_smiles)
+df.to_excel('metabolites_smiles.xlsx', index=False)
+
+
+
+filepath = 'D:\\All_Human_GTEx\\'
+df_model_mets = pd.read_excel(filepath+'metabolites.xlsx')
+# df_HMDB_mets = pd.read_excel(filepath+'副本hmdb_metabolites_217920_orign.xlsx')
+
+mets_id1 = df_model_mets['mets'].values.tolist()
+mets_id2 = df_model_mets['metsNoComp'].values.tolist()
+mets_SMILE = df_model_mets['SMILES'].values.tolist()
+
+mets_inchikey = []
+mets_inchi = []
+for i in mets_SMILE:
+    i = str(i)
+    if i == 'nan':
+        mets_inchikey.append('')
+        mets_inchi.append('')
+
+    elif i.startswith('*'):
+        mets_inchikey.append('')
+        mets_inchi.append('')
+
+    else:
+        # print(i)
+        mol = Chem.MolFromSmiles(i)
+        mets_inchikey.append(Chem.MolToInchiKey(mol))
+        mets_inchi.append(Chem.MolToInchi(mol))
+
+filepath1 = 'D:\\All_Human_GTEx\\'
+output = {'mets':mets_id1, 'metsNoComp':mets_id2, 'SMILES': mets_SMILE, 'inchikey':mets_inchikey, 'inchi':mets_inchi}
+output_tsv = pd.DataFrame(output)
+output_tsv.to_csv(filepath1+'metabolites_SMILES_Inchi.tsv', sep="\t", index=False)