Summary

• About
• Installation
• Documentation
• Usage

About

The Trident Chemwidgets (TCW) package provides a set of Jupyter widgets to improve data visibility in cheminformatics and molecular machine learning.

Trident Chemwidgets is a product of Trident Bioscience. Trident Bioscience builds machine learning applications for the biotech and pharmaceutical industries. If you're interested in learning more, you can visit us at trident.bio or email us at [email protected].

Installation

Trident Chemwidgets is available from pip and conda:

pip:

pip install trident-chemwidgets

conda:

conda install -c conda-forge trident-chemwidgets

Then, for Jupyter Lab, run:

jupyter labextension install trident-chemwidgets

Full instructions here.

Documentation

Click the link above to access the documentation. If you're looking for example notebooks, check out the links below (clone this repository to use them directly).

• JSME widget
• Histogram widget
• Scatter widget
• InteractiveMolecule widget

Usage

We suggest importing Trident Chemwidgets using the convention tcw, as shown below.

import trident_chemwidgets as tcw

Input functions

JSME

To facilitate molecular input, TCW provides an interface to the JSME Molecule Editor with added functionality.

editor = tcw.JSME()
editor

You can play with JSME widget in this notebook.

Plotting functions

Understanding the distribution of your data is critical to many cheminformatic tasks. To this end, we supply a number of interactive plotting functions that allow you to display both the data and underlying molecular structures. You can even subset and save the data based on the property values for use in downstream analysis.

Histogram

For 1-dimensional datasets, TCW provides a histogram plot function:

histogram = tcw.Histogram(data=data, smiles='smiles', x='tpsa', x_label='TPSA')
histogram

You can play with Histogram widget in this notebook.

Scatter plot

For 2-dimensional datasets, TCW provides a scatter plot function:

scatter = tcw.Scatter(data=data, smiles='smiles', x='mwt', y='logp',
                      x_label='Molecular Weight', y_label='logP')

You can play with Scatter widget in this notebook.

Output functions

Interactive molecule

To examine features at the atom level, TCW provides a function to plot a molecule and allow users to explore the atom-specific features:

mol = tcw.InteractiveMolecule('CC(C)CC1=CC=C(C=C1)C(C)C(=O)O', data=atom_data)
mol

You can play with InteractiveMolecule widget in this notebook.