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
« 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
9class AutomaticRulesExtraction:
10 """
11 A class for extracting automatic rules from chemical reactions.
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 """
25 def __init__(
26 self, existing_database: list = None, n_jobs: int = 4, verbose: int = 1
27 ) -> None:
28 """
29 Constructor for the class.
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
41 @staticmethod
42 def smiles_to_molecular_formula(smiles: str) -> str:
43 """
44 Convert a SMILES string to a molecular formula.
46 Args:
47 smiles (str): The SMILES string of the molecule.
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
57 def add_new_entries(self, filtered_fragments: dict) -> None:
58 """
59 Add new entries to the existing database.
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 )
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")
77 def extract_rules(self) -> List:
78 """
79 Extract automatic rules from the database.
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