Chemoface

Chemoface: Open-Source Software for Predicting Biological Activities

Open-source software for predicting biological activities of small molecules based on chemical structures, using machine learning models trained on compound datasets.

What is Chemoface?

Chemoface is an open-source computer program for predicting the biological activities of chemical compounds. It utilizes machine learning models that have been trained on large datasets of chemicals and their associated bioassay data to predict potential therapeutic effects and safety risks.

The key capabilities of Chemoface include:

• Predicting activity against a range of drug targets like enzymes, ion channels, and protein-protein interactions
• Estimating absorption, distribution, metabolism, excretion, and toxicity (ADMET) parameters
• Screening libraries of compounds to identify structures with desired bioactivities
• Prioritizing compounds for synthesis and testing
• Understanding structure-activity relationships

Some of the machine learning methods used by Chemoface include neural networks, random forests, support vector machines, and graph convolutions. It can make predictions based on simplified molecular-input line-entry system (SMILES) representations of compound structures.

Chemoface is implemented in Python and has an open-source MIT license allowing free usage. It provides both command line and graphical user interfaces. The models are customizable and new training data can be added to tailor predictions to specific drug discovery projects.

Chemoface Features

Official Links