A Simple Pharmacophore-Toolkit

The Pharmacophore-Toolkit is built on RDKit and allows for building simple pharmacophore models. The Pharmacophore-Toolset can generate models from crystal structures, docking poses, or SMILES string. To generate a 3D model, a .pml file will be generated. This files contains scripts to generate spheres with color and XYZ coordinates defined. The final 3D image can be rendered in PyMOL. Additional information can be found under the Tutorials section.

Install

You can install the Pharmacophore-Toolkit using pip:

pip install pharmacophore-toolkit

Alternatively, the environment can be created by cloning the repository and then running the following conda script:

conda env create -f environment.yaml

Note

Note: The Pharmacophore-Toolkit relies on cairosvg to create images before being converted into .png format. Depending on your workstation/machine, the CairoSVG package will need to be installed manually. Installation instructions can be found here. If it is not already installed on your machine globally, users can try conda to install cairosvg:

conda install conda-forge::cairosvg

Conda installation solved the problem on my machine, but may differ with yours.

Tutorials

Tutorials are written as JupyterNotebooks and can be found here. The Pharmacophore-Toolkit can generate several types of images:

Example Images

3D Model

A 3D conformation of molecules and alignment was performed using RDKit. Spheres can be generated for each query molecule to highlight pharmacophore features, such as Hydrogen Bond Donor/Acceptor, Hydrophobic, Aromatic, etc. The molecules can be generated in 3D using two methods:

• py3Dmol

The 3D model can be interactive in a Jupyter Notebook.

Images rendered in Jupyter Notebook using py3Dmol using a screenshot. Pharmacophore features for each molecule is highlighted. Blue spheres represent Hydrogen Bond Donors, gold spheres for Aromatic rings, and green for Hydrophobes.

• PyMOL

Images rendered in PyMOL using the generated .pml file. Molecules are colored using default settings and mirror those found using py3Dmol above. In this example, the spheres are mapped to Serotonin. Blue spheres represent Hydrogen Bond Donors and gold spheres represent Aromatic rings.

2D Model

The pharmacophore features can also be depicted as a 2D image, where each atom is highlighted with their respective pharmacophore features.

2D Pharmacophore images. Images were lightly edited to remove duplicate figure legends.

Similarity Map

For funsies, a method to generate Similarity Maps are also available. Final image generation requires Cairo installation. Additional information on how Similarity Maps work can be found on the RDKit blog.

Similarity map of molecules. All molecules were compared to Serotonin. More information can be seen at the RDKit documentation.