Coverage for SynRBL/SynRuleImputer/auto_extract_rules.py: 100%

26 statements  

« prev     ^ index     » next       coverage.py v7.4.3, created at 2024-04-11 11:17 +0200

1from rdkit import Chem 

2from rdkit.Chem import rdMolDescriptors 

3from joblib import Parallel, delayed 

4import pandas as pd 

5from typing import List 

6from SynRBL.SynRuleImputer.rule_data_manager import RuleImputeManager 

7 

8 

9class AutomaticRulesExtraction: 

10 """ 

11 A class for extracting automatic rules from chemical reactions. 

12 

13 Attributes 

14 ---------- 

15 rule_list : list 

16 A list of automatic rules extracted from reactions. 

17 existing_database : list 

18 A list of existing database entries. 

19 n_jobs : int, optional 

20 The number of jobs to run in parallel (default is -5). 

21 verbose : int, optional 

22 The verbosity level (default is 1). 

23 """ 

24 

25 def __init__( 

26 self, existing_database: list = None, n_jobs: int = 4, verbose: int = 1 

27 ) -> None: 

28 """ 

29 Constructor for the class. 

30 

31 Args: 

32 existing_database (list): A list representing an existing database. Default is an empty list. 

33 n_jobs (int): The number of jobs. Default is -5. 

34 verbose (int): The verbosity level. Default is 1. 

35 """ 

36 self.rule_list = [] 

37 self.existing_database = existing_database or [] 

38 self.n_jobs = n_jobs 

39 self.verbose = verbose 

40 

41 @staticmethod 

42 def smiles_to_molecular_formula(smiles: str) -> str: 

43 """ 

44 Convert a SMILES string to a molecular formula. 

45 

46 Args: 

47 smiles (str): The SMILES string of the molecule. 

48 

49 Returns: 

50 str: The molecular formula of the molecule. 

51 """ 

52 molecule = Chem.MolFromSmiles(smiles) 

53 if molecule: 

54 formula = rdMolDescriptors.CalcMolFormula(molecule) 

55 return formula 

56 

57 def add_new_entries(self, filtered_fragments: dict) -> None: 

58 """ 

59 Add new entries to the existing database. 

60 

61 Parameters 

62 ---------- 

63 filtered_fragments : dict 

64 A dictionary containing 'smiles' and 'formula' information for new entries. 

65 """ 

66 # Calculate molecular formulas in parallel 

67 molecular_formula = Parallel(n_jobs=self.n_jobs, verbose=self.verbose)( 

68 delayed(self.smiles_to_molecular_formula)(smi) 

69 for smi in filtered_fragments["smiles"] 

70 ) 

71 

72 # Create a DataFrame with 'smiles' and 'formula' and convert to a list of records 

73 self.new_smiles_dict = pd.DataFrame( 

74 {"formula": molecular_formula, "smiles": filtered_fragments["smiles"]} 

75 ).to_dict("records") 

76 

77 def extract_rules(self) -> List: 

78 """ 

79 Extract automatic rules from the database. 

80 

81 Returns 

82 ------- 

83 List 

84 A list of automatic rules extracted from the database. 

85 """ 

86 db = RuleImputeManager( 

87 self.existing_database 

88 ) # Create a RuleImputeManager instance 

89 db.add_entries(self.new_smiles_dict) # Add new entries to the database 

90 self.rule_list = ( 

91 db.database 

92 ) # Set self.rule_list to the database attribute of db 

93 return self.rule_list # Return self.rule_list