rdworks 0.39.1__tar.gz → 0.40.2__tar.gz

{rdworks-0.39.1 → rdworks-0.40.2}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: rdworks
-Version: 0.39.1
-Summary: Frequently used tasks built on RDKit and other tools
+Version: 0.40.2
+Summary: Routine tasks built on RDKit and other tools
 Author-email: Sung-Hun Bae <sunghun.bae@gmail.com>
 Maintainer-email: Sung-Hun Bae <sunghun.bae@gmail.com>
 License-Expression: MIT
@@ -10,7 +10,7 @@ Project-URL: Repository, https://github.com/sunghunbae/rdworks.git
 Project-URL: Issues, https://github.com/sunghunbae/rdworks/issues
 Project-URL: Changelog, https://github.com/sunghunbae/rdworks/blob/master/CHANGELOG.md
 Project-URL: Documentation, https://sunghunbae.github.io/rdworks/
-Keywords: cheminformatics,rdkit
+Keywords: cheminformatics,rdkit,cdpkit
 Classifier: Development Status :: 3 - Alpha
 Classifier: Intended Audience :: Developers
 Classifier: Programming Language :: Python :: 3
@@ -30,5 +30,45 @@ Requires-Dist: bitarray
 Requires-Dist: pytest
 Dynamic: license-file
 
-# rdworks
-Higher level wrapper using RDKit
+# Rdworks - routine tasks made easy
+
+Rdworks is designed to perform routine cheminformatics tasks easily. It is built on RDKit and other tools. 
+
+## Install
+
+```sh
+$ pip install rdworks
+```
+
+OR
+
+```sh
+$ pip install git+https://github.com/sunghunbae/rdworks.git
+```
+
+## Getting started
+
+```py
+from rdworks import Mol
+
+version = rdworks.__version__
+
+mol = Mol('CC(=O)Nc1ccc(O)cc1', 'acetaminophen')
+
+mol = mol.make_confs(n=5)
+mol.to_sdf('acetaminophen.sdf')
+
+torsion_dict = mol.torsion_atoms() 
+# torsion_dict = {0: (5,4,3,1)}
+
+mol = mol.torsion_energies(calculator='MMFF94', simplify=True)
+
+mol.plot_torsion_energies(0, figsize=(6,4))
+
+mol.to_png(300, 300, atom_index=True, highlight_atoms=torsion_dict.get(0))
+
+serialized = mol.serialize(compress=True)
+mol2 = Mol().deserialize(serialized, compress=True)
+
+mol3 = mol.copy()
+```

rdworks-0.40.2/README.md

@@ -0,0 +1,42 @@
+# Rdworks - routine tasks made easy
+
+Rdworks is designed to perform routine cheminformatics tasks easily. It is built on RDKit and other tools. 
+
+## Install
+
+```sh
+$ pip install rdworks
+```
+
+OR
+
+```sh
+$ pip install git+https://github.com/sunghunbae/rdworks.git
+```
+
+## Getting started
+
+```py
+from rdworks import Mol
+
+version = rdworks.__version__
+
+mol = Mol('CC(=O)Nc1ccc(O)cc1', 'acetaminophen')
+
+mol = mol.make_confs(n=5)
+mol.to_sdf('acetaminophen.sdf')
+
+torsion_dict = mol.torsion_atoms() 
+# torsion_dict = {0: (5,4,3,1)}
+
+mol = mol.torsion_energies(calculator='MMFF94', simplify=True)
+
+mol.plot_torsion_energies(0, figsize=(6,4))
+
+mol.to_png(300, 300, atom_index=True, highlight_atoms=torsion_dict.get(0))
+
+serialized = mol.serialize(compress=True)
+mol2 = Mol().deserialize(serialized, compress=True)
+
+mol3 = mol.copy()
+```

{rdworks-0.39.1 → rdworks-0.40.2}/pyproject.toml

@@ -14,17 +14,17 @@ dependencies = [
     "tqdm",
     "ase",
     # "xtb<6.3.0", <--- installation with conda, issue with FORTRAN runtime error
-    "mendeleev",
+    "mendeleev", # periodic table
     "scour", # SVG optimization
     "bitarray", # dependencies of bitqt
     "pytest",
     ]
 authors = [{name = "Sung-Hun Bae", email="sunghun.bae@gmail.com"}, ]
 maintainers = [{name = "Sung-Hun Bae", email="sunghun.bae@gmail.com"}, ]
-description = "Frequently used tasks built on RDKit and other tools"
+description = "Routine tasks built on RDKit and other tools"
 readme = "README.md"
 license = "MIT"
-keywords = ["cheminformatics", "rdkit",]
+keywords = ["cheminformatics", "rdkit", "cdpkit"]
 classifiers = [
     "Development Status :: 3 - Alpha",
     "Intended Audience :: Developers",

{rdworks-0.39.1 → rdworks-0.40.2}/src/rdworks/__init__.py

@@ -1,4 +1,4 @@
-__version__ = '0.39.1'
+__version__ = '0.40.2'
 
 from rdworks.conf    import Conf
 from rdworks.mol     import Mol

{rdworks-0.39.1 → rdworks-0.40.2}/src/rdworks/conf.py

@@ -19,7 +19,7 @@ from PIL import Image
 from rdworks.std import clean_2d
 from rdworks.torsion import get_torsion_atoms, create_torsion_fragment
 from rdworks.units import ev2kcalpermol, pm2angstrom
-from rdworks.utils import recursive_round
+from rdworks.utils import recursive_round, compress_string, decompress_string
 from rdworks.xtb.wrapper import GFN2xTB
 from rdworks.display import render_png, render_svg
 
@@ -60,7 +60,10 @@ class Conf:
 
         if isinstance(molecule, str): # 3-D MolBLock string
             try:
-                self.rdmol = Chem.MolFromMolBlock(molecule)
+                self.rdmol = Chem.MolFromMolBlock(molecule, 
+                                                  sanitize=False, 
+                                                  removeHs=False, 
+                                                  strictParsing=True)
             except:
                 ValueError(f'Conf() Error: invalid MolBlock string')
 
@@ -581,7 +584,7 @@ class Conf:
             return json.dumps(self.props)
     
 
-    def serialize(self, decimals: int = 3) -> str:
+    def serialize(self, decimals: int = 3, compressed: bool = False) -> str:
         """Serialize information necessary to rebuild a Conf object.
         Args:
             decimals (int, optional): number of decimal places for float data type. Defaults to 3.
@@ -598,10 +601,13 @@ class Conf:
             'molblock' : self.to_molblock(),
             })
 
+        if compressed:
+            serialized = compress_string(serialized)
+            
         return serialized
 
 
-    def deserialize(self, serialized: str) -> Self:
+    def deserialize(self, serialized: str, compressed: bool = False) -> Self:
         """De-serialize information and rebuild a Conf object.
         
         Example:
@@ -614,6 +620,9 @@ class Conf:
         Returns:
             Self: modified self.
         """
+        if compressed:
+            serialized = decompress_string(serialized)
+
         data = json.loads(serialized)
 
         self.name = data['name']
@@ -626,9 +635,12 @@ class Conf:
         return self
 
 
-    def to_molblock(self) -> str:
+    def to_molblock(self, compressed: bool = False) -> str:
         """Returns MolBlock"""
-        return Chem.MolToMolBlock(self.rdmol)
+        molblock = Chem.MolToMolBlock(self.rdmol)
+        if compressed:
+            molblock = compress_string(molblock)
+        return molblock
 
 
     def to_xyz(self) -> str:

{rdworks-0.39.1 → rdworks-0.40.2}/src/rdworks/mol.py

@@ -37,7 +37,7 @@ from rdworks.std import generate_inchi_key, desalt_smiles, standardize, clean_2d
 from rdworks.xml import list_predefined_xml, get_predefined_xml, parse_xml
 from rdworks.scaffold import rigid_fragment_indices
 from rdworks.descriptor import rd_descriptor, rd_descriptor_f
-from rdworks.utils import convert_tril_to_symm, QT, recursive_round
+from rdworks.utils import convert_tril_to_symm, QT, recursive_round, compress_string, decompress_string
 from rdworks.units import ev2kcalpermol
 from rdworks.autograph import NMRCLUST, DynamicTreeCut, RCKmeans, AutoGraph
 from rdworks.bitqt import BitQT
@@ -1558,7 +1558,7 @@ class Mol:
         return json.dumps(props)
 
 
-    def serialize(self, decimals: int = 2) -> str:
+    def serialize(self, decimals: int = 2, compressed: bool = False) -> str:
         """Serialize information necessary to rebuild a Mol object.
 
         Args:
@@ -1575,10 +1575,13 @@ class Mol:
             'confs' : [conf.serialize() for conf in self.confs],
             })
         
+        if compressed:
+            serialized = compress_string(serialized)
+
         return serialized
 
     
-    def deserialize(self, serialized: str) -> Self:
+    def deserialize(self, serialized: str, compressed: bool = False) -> Self:
         """De-serialize the information and build a new Mol object.
 
         Example:
@@ -1591,6 +1594,9 @@ class Mol:
         Returns:
             Self: modified self.
         """
+        if compressed:
+            serialized = decompress_string(serialized)
+            
         data = json.loads(serialized)
 
         self.name = data['name']

rdworks-0.40.2/src/rdworks/testdata.py

@@ -0,0 +1,24 @@
+drugs = {
+    'Acetaminophen': 'CC(=O)Nc1ccc(O)cc1',
+    'Aspirin': 'CC(=O)OC1=CC=CC=C1C(=O)O',
+    'Atorvastatin': 'CC(C)C1=C(C(=C(N1CC[C@H](C[C@H](CC(=O)O)O)O)C2=CC=C(C=C2)F)C3=CC=CC=C3)C(=O)NC4=CC=CC=C4',
+    'Atovaquone': 'C1CC(CCC1C2=CC=C(C=C2)Cl)C3=C(C4=CC=CC=C4C(=O)C3=O)O',
+    'Cefdinir': 'C=CC1=C(N2[C@@H]([C@@H](C2=O)NC(=O)/C(=N\\O)/C3=CSC(=N3)N)SC1)C(=O)O',
+    'Chlorprothixene': 'CN(CC/C=C1C2=CC=CC=C2SC3=C/1C=C(Cl)C=C3)C',
+    'Cimetidine': 'CC1=C(N=CN1)CSCCNC(=NC)NC#N',
+    'Clomipramine': 'CN(C)CCCN1C2=CC=CC=C2CCC3=C1C=C(C=C3)Cl',
+    'Ethopropazine': 'CCN(CC)C(C)CN1C2=CC=CC=C2SC3=CC=CC=C31',
+    'Famotidine': 'C1=C(N=C(S1)N=C(N)N)CSCC/C(=N/S(=O)(=O)N)/N',
+    'Fluconazole': 'C1=CC(=C(C=C1F)F)C(CN2C=NC=N2)(CN3C=NC=N3)O',
+    'Granisetron': 'CN1[C@@H]2CCC[C@H]1CC(C2)NC(=O)C3=NN(C4=CC=CC=C43)C',
+    'Leflunomide': 'CC1=C(C=NO1)C(=O)NC2=CC=C(C=C2)C(F)(F)F',
+    'Linezolid': 'CC(=O)NC[C@H]1CN(C(=O)O1)C2=CC(=C(C=C2)N3CCOCC3)F',
+    'Methixene': 'CN1CCCC(C1)CC2C3=CC=CC=C3SC4=CC=CC=C24',
+    'Molindone': 'CCC1=C(NC2=C1C(=O)C(CC2)CN3CCOCC3)C',
+    'Paroxetine': 'C1CNC[C@H]([C@@H]1C2=CC=C(C=C2)F)COC3=CC4=C(C=C3)OCO4',
+    'Pergolide': 'CCCN1C[C@@H](C[C@H]2[C@H]1CC3=CNC4=CC=CC2=C34)CSC',
+    'Rifampin': 'C[C@H]1/C=C/C=C(\\C(=O)NC2=C(C(=C3C(=C2O)C(=C(C4=C3C(=O)[C@](O4)(O/C=C/[C@@H]([C@H]([C@H]([C@@H]([C@@H]([C@@H]([C@H]1O)C)O)C)OC(=O)C)C)OC)C)C)O)O)/C=N/N5CCN(CC5)C)/C',
+    'Simvastatin': 'O=C(O[C@@H]1[C@H]3C(=C/[C@H](C)C1)\\C=C/[C@@H]([C@@H]3CC[C@H]2OC(=O)C[C@H](O)C2)C)C(C)(C)CC',
+    'Sitagliptin': 'Fc1cc(c(F)cc1F)C[C@@H](N)CC(=O)N3Cc2nnc(n2CC3)C(F)(F)F',
+    'Sofosbuvir': 'C[C@@H](C(OC(C)C)=O)N[P@](OC[C@@H]1[C@H]([C@@](F)([C@@H](O1)N2C=CC(NC2=O)=O)C)O)(OC3=CC=CC=C3)=O',
+    }

{rdworks-0.39.1 → rdworks-0.40.2}/src/rdworks/utils.py

@@ -1,6 +1,8 @@
 import numpy as np
 import math
 import operator
+import gzip
+import base64
 
 from pathlib import Path
 from types import SimpleNamespace
@@ -14,6 +16,33 @@ from rdkit import Chem
 from rdworks.autograph.centroid import centroid_medoid
 
 
+def compress_string(text: str) -> str:
+    """compress string to base64-encoded string.
+
+    Args:
+        text (str): original string.
+
+    Returns:
+        str: base64-encoded compressed string.
+    """
+    compressed_bytes = gzip.compress(text.encode('utf-8'))
+    return base64.b64encode(compressed_bytes).decode('utf-8')
+
+
+def decompress_string(compressed_text: str) -> str:
+    """decompress base64-encoded string to original string.
+
+    Args:
+        compressed_text (str): base64-encoded compressed string.
+
+    Returns:
+        str: original string.
+    """
+    compressed_bytes = base64.b64decode(compressed_text)
+    decompressed_text = gzip.decompress(compressed_bytes).decode('utf-8')
+    return decompressed_text
+
+
 def compute(fn:Callable, largs: list, **kwargs) -> list:
     max_workers = kwargs.get('max_workers', 1)
     chunksize   = kwargs.get('chunksize', 10)

{rdworks-0.39.1 → rdworks-0.40.2}/src/rdworks.egg-info/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: rdworks
-Version: 0.39.1
-Summary: Frequently used tasks built on RDKit and other tools
+Version: 0.40.2
+Summary: Routine tasks built on RDKit and other tools
 Author-email: Sung-Hun Bae <sunghun.bae@gmail.com>
 Maintainer-email: Sung-Hun Bae <sunghun.bae@gmail.com>
 License-Expression: MIT
@@ -10,7 +10,7 @@ Project-URL: Repository, https://github.com/sunghunbae/rdworks.git
 Project-URL: Issues, https://github.com/sunghunbae/rdworks/issues
 Project-URL: Changelog, https://github.com/sunghunbae/rdworks/blob/master/CHANGELOG.md
 Project-URL: Documentation, https://sunghunbae.github.io/rdworks/
-Keywords: cheminformatics,rdkit
+Keywords: cheminformatics,rdkit,cdpkit
 Classifier: Development Status :: 3 - Alpha
 Classifier: Intended Audience :: Developers
 Classifier: Programming Language :: Python :: 3
@@ -30,5 +30,45 @@ Requires-Dist: bitarray
 Requires-Dist: pytest
 Dynamic: license-file
 
-# rdworks
-Higher level wrapper using RDKit
+# Rdworks - routine tasks made easy
+
+Rdworks is designed to perform routine cheminformatics tasks easily. It is built on RDKit and other tools. 
+
+## Install
+
+```sh
+$ pip install rdworks
+```
+
+OR
+
+```sh
+$ pip install git+https://github.com/sunghunbae/rdworks.git
+```
+
+## Getting started
+
+```py
+from rdworks import Mol
+
+version = rdworks.__version__
+
+mol = Mol('CC(=O)Nc1ccc(O)cc1', 'acetaminophen')
+
+mol = mol.make_confs(n=5)
+mol.to_sdf('acetaminophen.sdf')
+
+torsion_dict = mol.torsion_atoms() 
+# torsion_dict = {0: (5,4,3,1)}
+
+mol = mol.torsion_energies(calculator='MMFF94', simplify=True)
+
+mol.plot_torsion_energies(0, figsize=(6,4))
+
+mol.to_png(300, 300, atom_index=True, highlight_atoms=torsion_dict.get(0))
+
+serialized = mol.serialize(compress=True)
+mol2 = Mol().deserialize(serialized, compress=True)
+
+mol3 = mol.copy()
+```

{rdworks-0.39.1 → rdworks-0.40.2}/src/rdworks.egg-info/SOURCES.txt

@@ -16,6 +16,7 @@ src/rdworks/scaffold.py
 src/rdworks/std.py
 src/rdworks/stereoisomers.py
 src/rdworks/tautomers.py
+src/rdworks/testdata.py
 src/rdworks/torsion.py
 src/rdworks/units.py
 src/rdworks/utils.py

{rdworks-0.39.1 → rdworks-0.40.2}/tests/test_basics.py

@@ -511,6 +511,7 @@ def test_torsion_fragment_from_conf():
     assert frag_ijkl == ta2[0]
 
     ref_conf2 = ref_conf2.torsion_energies(calculator='MMFF94', interval=15)
+    ref_conf3 = ref_conf2.torsion_energies_one(calculator='MMFF94', indices=frag_ijkl)
 
 
 def test_torsion_energies():
@@ -549,4 +550,4 @@ def test_serialization():
     rebuilt = Mol().deserialize(serialized)
     assert rebuilt.count() == 10
     assert rebuilt.name == name
-    assert rebuilt == mol
+    assert rebuilt == mol

rdworks-0.39.1/README.md

@@ -1,2 +0,0 @@
-# rdworks
-Higher level wrapper using RDKit