Allowing multiple solutions to be returned in order to use weights for RMSD/overlap

.github/workflows/CI.yml

@@ -44,7 +44,7 @@ jobs:
       - name: Install ties
         shell: bash -l {0}
         run: |
-          python -m pip install . --no-deps -v
+          python -m pip install --no-deps . 
 
       - name: Install pytest
         shell: bash -l {0}

environment.yml

@@ -4,9 +4,9 @@ channels:
   - defaults
 
 dependencies:
-  - conda-forge::ambertools >=20.0
+  - conda-forge::ambertools >=21.0
   - ties_md
-  - numpy <=1.23.5
+  - numpy
   - cython
   - setuptools
   - matplotlib

pyproject.toml

@@ -0,0 +1,3 @@
+[build-system]
+requires = ["setuptools"]
+build-backend = "setuptools.build_meta"

setup.py

@@ -1,14 +1,11 @@
 from setuptools import setup, find_packages
 from distutils.extension import Extension
-from pathlib import Path
 
-import numpy
-
-# get the information about numpy header files
 try:
-    numpy_include = numpy.get_include()
-except AttributeError:
-    numpy_include = numpy.get_numpy_include()
+    import numpy
+    numpy.get_include()
+except ModuleNotFoundError:
+    pass
 
 # handle cython modules: pyqcprot module for the rotation matrix
 try:
@@ -22,7 +19,7 @@
     print (f'use_cython: {use_cython}')
 
 ext_modules = [Extension("ties/pyqcprot_ext/pyqcprot", [f"ties/pyqcprot_ext/pyqcprot.{'pyx' if use_cython else 'c'}"],
-                         include_dirs=[numpy_include],
+                         include_dirs=[],
                          extra_compile_args=["-O3","-ffast-math"])]
 
 setup(
@@ -35,15 +32,6 @@
     author_email='[email protected]',
     packages=find_packages(),
     include_package_data=True,
-    install_requires=[
-        'numpy',
-        'cython',
-        'setuptools',
-        'matplotlib',
-        'networkx',
-        'dimod',
-        'tabulate',
-        'dwave-networkx'],
     entry_points={
         'console_scripts': [
             'ties = ties:cli.command_line_script'

ties/cli.py

@@ -30,17 +30,17 @@ def command_line_script():
                              'If more than 2 ligands are provided, '
                              'Lead Optimisation Mapping (TIES MAP) will be used.')
     parser.add_argument('-dir', '--ties-output-dir', metavar='directory', dest='workdir',
-                        type=pathlib.Path, required=False,
-                        help='If not provided, "ties20" directory will be created in the current directory. ')
+                        type=pathlib.Path, required=False, default="ties",
+                        help='The destination directory for the output. Default: "ties".')
     parser.add_argument('-nc', '--ligand-net-charge', metavar='integer', dest='ligand_net_charge',
                         type=int, required=False,
                         help='An integer representing the net charge of each ligand. '
-                             'All ligands must have the same charge.')
+                             'All ligands must have the same charge. Default: 0.1e.')
     parser.add_argument('-p', '--protein', metavar='file', dest='protein',
                         type=ArgparseChecker.existing_file, required=False,
-                        help='The protein file')
+                        help='The protein file. Not necessary. ')
     parser.add_argument('-qtol', '--q-pair-tolerance', metavar='decimal', dest='atom_pair_q_atol',
-                        type=float, required=False, default=0.1,
+                        type=float, required=False,
                         help='The maximum difference in charge between any two paired atoms (electron unit). '
                              'Default 0.1e')
     parser.add_argument('-netqtol', '--q-net-tolerance', metavar='decimal', dest='net_charge_threshold',
@@ -133,6 +133,12 @@ def command_line_script():
     parser.add_argument('-elements', '--compare-elements', metavar='boolean', dest='use_element',
                         type=ArgparseChecker.str2bool, required=False, default=False,
                         help='Ignore the specific atom types in the superimposition. Use only the elements. ')
+    parser.add_argument('-weights', '--mcs-rmsd-weights', metavar='str', dest='weights_ratio',
+                        type=ArgparseChecker.ratio, required=False,
+                        help='The weights for the weighted sum of 1) MCS overlap size to 2) RMSD '
+                             'when coordinates are used for selection of the best structure. '
+                             'Default it is "1:1" for "MCS:RMSD".  '
+                             'MCS is defined (1 - MCS fraction), so lower value is better.')
 
     # initialise the config class
     args = parser.parse_args()

ties/config.py

@@ -46,6 +46,8 @@ def __init__(self, **kwargs):
         # use only the element in the superimposition rather than the specific atom type
         self._use_element = False
         self._use_element_in_superimposition = True
+        # weights in choosing the best MCS, the weighted sum of "(1 - MCS fraction) and RMSD".
+        self.weights = [1, 0.5]
 
         # coordinates
         self._align_molecules_using_mcs = False
@@ -82,8 +84,7 @@ def __init__(self, **kwargs):
         self.uses_cmd = False
 
         # assign all the initial configuration values
-        for key, value in kwargs.items():
-            setattr(self, key, value)
+        self.set_configs(**kwargs)
 
     @property
     def workdir(self):
@@ -897,4 +898,8 @@ def get_serializable(self):
     def set_configs(self, **kwargs):
         # set all the configs one by one
         for k,v in kwargs.items():
+            # skip None values, these come from the command line
+            if v is None:
+                continue
+
             setattr(self, k, v)

ties/helpers.py

@@ -124,6 +124,13 @@ def str2bool(v):
         else:
             raise argparse.ArgumentTypeError('Boolean value expected.')
 
+    @staticmethod
+    def ratio(v):
+        if ':' not in v:
+            argparse.ArgumentTypeError('The ratio has to be separated by ":".')
+        mcs, rmsd = v.split(':')
+        return [float(mcs), float(rmsd)]
+
     @staticmethod
     def existing_file(v):
         # check ligand arguments

ties/pair.py

@@ -155,7 +155,7 @@ def superimpose(self, **kwargs):
             print('Starting nodes will be used:', starting_node_pairs)
 
         # fixme - simplify to only take the ParmEd as input
-        suptops = superimpose_topologies(ligand1_nodes.values(), ligand2_nodes.values(),
+        suptop = superimpose_topologies(ligand1_nodes.values(), ligand2_nodes.values(),
                                          disjoint_components=self.config.allow_disjoint_components,
                                          net_charge_filter=True,
                                          pair_charge_atol=self.config.atom_pair_q_atol,
@@ -173,16 +173,16 @@ def superimpose(self, **kwargs):
                                          force_mismatch=new_mismatch_names,
                                          starting_node_pairs=starting_node_pairs,
                                          parmed_ligA=parmed_ligA, parmed_ligZ=parmed_ligZ,
-                                         starting_pair_seed=self.config.superimposition_starting_pair)
+                                         starting_pair_seed=self.config.superimposition_starting_pair,
+                                         config=self.config)
 
-        assert len(suptops) == 1
-        self.set_suptop(suptops[0], parmed_ligA, parmed_ligZ)
+        self.set_suptop(suptop, parmed_ligA, parmed_ligZ)
         # attach the used config to the suptop
-        suptops[0].config = self.config
+        suptop.config = self.config
         # attach the morph to the suptop
-        suptops[0].morph = self
+        suptop.morph = self
 
-        return suptops[0]
+        return suptop
 
     def set_suptop(self, suptop, parmed_ligA, parmed_ligZ):
         """