Molecular deep learning at the edge of chemical space

Derek van Tilborg, Luke Rossen, Francesca Grisoni
Corresponding author: [email protected]

Abstract
Bla bla

Figure 1. The architecture of the Joint Molecular Model (JMM) estimates how ‘unfamiliar’ a molecule is to the model through its reconstruction loss.

Prerequisites

The following Python packages are required to run this codebase. Tested on macOS 15.1.1

• PyTorch (1.12.1)
• Pandas (1.5.3)
• Numpy (1.23.5)
• XGBoost (1.7.3)
• Scikit-learn (1.2.1)

Installation

Install dependencies from the provided env.yaml file. This typically takes a couple of minutes.

conda env create -f env.yaml

Content

This repository is structured in the following way:

• data: contains all data
• cheminformatics: the starting data set
• experiments: all Python scripts required to replicate the study
• jcm: all deep learning code
• results: collection of results
• plots: all scripts required to plot the figures in the paper

How to cite

You can currently cite our pre-print:

van Tilborg et al. (2025). Molecular deep learning at the edge of chemical space. ChemRxiv.

License

This codebase is under MIT license. For use of specific models, please refer to the model licenses found in the original packages.