feat: Added usage of SMILES for the main functions

README.md

@@ -51,6 +51,16 @@ M  END
 """
 
 std_molblock = standardizer.standardize_molblock(o_molblock)
+```
+> **Note**
+> Using SMILES:
+
+```python
+from chembl_structure_pipeline import standardizer
+
+SMILES = "CN1C=NC2=C1C(=O)N(C(=O)N2C)C"
+std_smiles = standardizer.standardize_molblock_from_smiles(SMILES, get_smiles=True)
+
 ```
 
 ### Get the parent compound [(info)](https://github.com/chembl/ChEMBL_Structure_Pipeline/wiki/Work-done-by-each-step#get_parent_molblock)
@@ -74,6 +84,17 @@ M  END
 parent_molblock, _ = standardizer.get_parent_molblock(o_molblock)
 ```
 
+> **Note**
+> Using SMILES:
+
+```python
+from chembl_structure_pipeline import standardizer
+
+SMILES = "CN1C=NC2=C1C(=O)N(C(=O)N2C)C"
+parent_smiles = standardizer.get_parent_mol_from_smiles(SMILES, get_smiles=True)
+
+```
+
 ### Check a compound [(info)](https://github.com/chembl/ChEMBL_Structure_Pipeline/wiki/Work-done-by-each-step#checkmolecule)
 
 The checker assesses the quality of a structure. It highlights specific features or issues in the structure that may need to be revised. Together with the description of the issue, the checker process returns a penalty score (between 0-9) which reflects the seriousness of the issue (the higher the score, the more critical is the issue)
@@ -96,6 +117,16 @@ M  END
 """
 
 issues = checker.check_molblock(o_molblock)
+```
+> **Note**
+> Using SMILES:
+
+```python
+from chembl_structure_pipeline import checker
+
+SMILES = "CN1C=NC2=C1C(=O)N(C(=O)N2C)C"
+issues = checker.check_mol_from_smiles(SMILES)
+
 ```
 
 ## References

chembl_structure_pipeline/__init__.py

@@ -101,7 +101,7 @@
 from .standardizer import standardize_molblock, standardize_mol
 from .standardizer import get_parent_molblock, get_parent_mol
 
-__version__ = "1.2.0"
+__version__ = "1.2.1"
 
 #
 #  Copyright (c) 2019 Greg Landrum

chembl_structure_pipeline/checker.py

@@ -508,3 +508,17 @@ def check_molblock(mb):
     if tpl:
         res.append(tpl)
     return tuple(sorted(res, reverse=True))
+
+
+def check_mol_from_smiles(smiles: str, sanitize=False):
+    """
+    Use the check_molblock function to determine
+    if the molecule has any issues based on
+    a given SMILES string.
+    Args (str): SMILES string.
+    Returns (tuple): Issues with smiles
+    """
+    mol = Chem.MolFromSmiles(smiles, sanitize)
+    if mol:
+        mol_block = Chem.MolToMolBlock(mol)
+        return check_molblock(mol_block)

chembl_structure_pipeline/standardizer.py

@@ -453,6 +453,26 @@ def get_parent_molblock(ctab, neutralize=True, check_exclusion=True, verbose=Fal
     return Chem.MolToMolBlock(parent, kekulize=False), exclude
 
 
+def get_parent_mol_from_smiles(
+    smiles: str, get_smiles: bool = False, sanitize: bool = False
+):
+    """
+    Use the get_parent_molblock function to retraive the parent molblock
+    using the given SMILES string
+    Args (str): SMILES string.
+    Returns (mol): parent mol.
+    """
+    mol = Chem.MolFromSmiles(smiles, sanitize)
+    if mol:
+        mol_block = Chem.MolToMolBlock(mol)
+        if get_smiles:
+            parent_molblock,_ = get_parent_molblock(mol_block)
+            rdkit_mol = Chem.MolFromMolBlock(parent_molblock)
+            parentsmiles = Chem.MolToSmiles(rdkit_mol)
+            return parentsmiles
+        return get_parent_molblock(mol_block)
+
+
 def standardize_mol(m, check_exclusion=True, sanitize=True):
     if check_exclusion:
         exclude = exclude_flag(m, includeRDKitSanitization=False)
@@ -512,3 +532,23 @@ def standardize_molblock(ctab, check_exclusion=True):
         if exclude_flag(m, includeRDKitSanitization=False):
             return ctab
     return Chem.MolToMolBlock(standardize_mol(m, check_exclusion=False, sanitize=False))
+
+
+def standardize_molblock_from_smiles(
+    smiles: str, get_smiles: bool = False, sanitize: bool = False
+):
+    """
+    Use the standardize_molblock function to identify issues and fix it for
+    a molecule based on the given SMILES string.
+    Args (str): SMILES string.
+    Returns (mol): Fixed molecule.
+    """
+    mol = Chem.MolFromSmiles(smiles, sanitize)
+    if mol:
+        mol_block = Chem.MolToMolBlock(mol)
+        if get_smiles:
+            standardized_mol = standardize_molblock(mol_block)
+            rdkit_mol = Chem.MolFromMolBlock(standardized_mol)
+            standardizedsmiles = Chem.MolToSmiles(rdkit_mol)
+            return standardizedsmiles
+        return standardize_molblock(mol_block)