This package includes various tools that is being used on the ongoing computational modelling studies of macrocycles, with particular keen on cyclic peptidic species.
The functionalities are primarily split into three submodules:
- Molgen : to construct an insillico (cyclic)molecular entity that contains the atom names (for atom type assignment) and connectivity information including bond orders.
- Confgen : to construct 3D coordinates for atoms in a molecular entity
- Metrics : to analyse a given set of 3D structures
The tools rely heavily on RDKit and mdtraj functionalities and enable interconversions between these two molecular representations.
This package only supports Python 3.5 or above.
The cPeptools packages is built on various packages that is conda-installable. Therefore, either anaconda or miniconda is needed as prerequisites.
Once conda is available, execute the following commands to install all essential dependencies:
conda install -c anaconda numpy scipy scikit-learn
conda install -c omnia mdtraj
conda install -c rdkit rdkit
Please ensure your rdkit version is v.2020.03 or above.
After all dependencies are installed, navigate to a directory where you wish to clone this repository and execute:
git clone https://github.com/rinikerlab/cpeptools.git
cd cpeptools
python setup.py build install
Published use involving cPeptools is documented here.
In the method section of this publication, we describe the use of elliptical 2D geometry to confine the conformational space of conformer generation. Figure 2 is repreduced below, and we would like to bias sampling of the rightmost-like cyclic molecules. This requires the useage of this package.
The following code snippet is the minimum code needed:
import cpeptools
from rdkit import Chem
from rdkit.Chem import AllChem
mol = Chem.MolFromSmiles("C-C(-C)-C-[C@@H]1-N-C(=O)-[C@H](-C)-N(-C)-C(=O)-[C@H]2-C-C-C-N-2-C(=O)-[C@H](-C-C(-C)-C)-N-C(=O)-[C@H](C)-N(-C)-C(=O)-[C@H](-C-C(-C)-C)-N-C(=O)-[C@H](-C)-N(-C)-C(=O)-[C@H]2-C-C-C-N-2-C(=O)-[C@H](-C-C(-C)-C)-N-C(=O)-[C@H](-C)-N(-C)-C-1=O")
mol = Chem.AddHs(mol)
params = AllChem.ETKDGv3() #contains the modified macrocycle torsion workflow
bmat = cpeptools.bound_matrix_from_ellipse(mol, angle = 0, eccentricity = 0.99)
params.SetBoundsMat(bmat)
AllChem.EmbedMultipleConfs(mol, 3 , params) #conformers written to mol objectCheck the documentation for bound_matrix_from_ellipse for details of additional parameters.
The usage of custom pairwise Coulombic interactions (CPCIs) is also described in the method section. The following code gets CPCIS specific for amide interactions:
amide_cpci = cpeptools.get_amide_pairwise_coulomb_interaction(mol)
params.SetCPCI(amide_cpci)
AllChem.EmbedMultipleConfs(mol, 3 , params) #conformers written to mol objectCheck the documentation for get_amide_pairwise_coulomb_interaction for details of additional parameters.
Another useage described in the method section is to map a 3D ring structure to a 2D represenation with calculated metrics, figure 3 of the paper is reproduced below:
The following code depicts a 3D conformer (left) as a 2D representation (right) with calculated metrics:
from cpeptools import metrics
import mdtraj as md
smiles = "C-C(-C)-C-[C@@H]1-N-C(=O)-[C@H](-C)-N(-C)-C(=O)-[C@H]2-C-C-C-N-2-C(=O)-[C@H](-C-C(-C)-C)-N-C(=O)-[C@H](C)-N(-C)-C(=O)-[C@H](-C-C(-C)-C)-N-C(=O)-[C@H](-C)-N(-C)-C(=O)-[C@H]2-C-C-C-N-2-C(=O)-[C@H](-C-C(-C)-C)-N-C(=O)-[C@H](-C)-N(-C)-C-1=O"
metrics.show_ellipse_fitting("./examples/data/Decapeptide.pdb", smiles = smiles)See doc for show_ellipse_fitting for options for additional parameters. To only calculate the values without drawing the figure see calculate_eccentricity.
To cite this work:
@article{Wang2020,
doi = {10.1021/acs.jcim.0c00025},
url = {https://doi.org/10.1021/acs.jcim.0c00025},
year = {2020},
month = mar,
publisher = {American Chemical Society ({ACS})},
author = {Shuzhe Wang and Jagna Witek and Gregory A. Landrum and Sereina Riniker},
title = {Improving Conformer Generation for Small Rings and Macrocycles Based on Distance Geometry and Experimental Torsional-Angle Preferences},
journal = {Journal of Chemical Information and Modeling}
}
Copyright (c) 2018, shuzhe Wang
Project based on the Computational Molecular Science Python Cookiecutter