Interactive Similarity Network using python

Script developed to build an interactive molecular similarity network to visualize Tanimoto similarity between molecules in a dataset.

Libraries used

• pandas - a Python package that provides fast, flexible, and expressive data structures designed to make working with "relational" or "labeled" data both easy and intuitive.

• NumPy - the fundamental package for array computing with Python

• RDKit - Open source toolkit for cheminformatics

• NetworkX - a Python package for the creation, manipulation, and study of the structure, dynamics, and functions of complex networks.

• Matplotlib - a comprehensive library for creating static, animated, and interactive visualizations in Python.

• pyvis - Interactive network visualizations.

Libraries were used in a Miniconda3 environment using python 3.6.13

Instalation

Miniconda3: Installation

pandas:

conda install -c anaconda pandas

numpy

conda install -c anaconda numpy

RDKit

conda install -c rdkit rdkit

NetworkX

conda install -c anaconda networkx

Matplotlib

conda install -c conda-forge matplotlib

pyvis

conda install -c conda-forge pyvis

How to run

Download the code and unzip it on the desirable directory

To run use the following command:

python similarityNetwork.py

• The dataset layout should be in the format 'smiles' 'molecule_name' as presented in the dataset_ds.smi

• The threshold may be changed in line 40 as it follows

Tc = DataStructs.TanimotoSimilarity(fps[i], fps[j])
        **if Tc >= 0.3:**
            g.add_edge(smiles[i], smiles[j], length=1000)

Observations

This script has been elaborated using as references the following articles and codes:

• Draw molecular network on Jupyter notebook with rdkit and cytoscape.js - code

• Molecular similarity network with visualised structures - code

Authorship

• Author: Brenda Ferrari (brendaferrari)
• Co-author: Camilo Lima (limacamilo)

Social preview original photo by Brenda Ferrari (brendaferrari)