Abstract:Abstract Generative models have increasingly been proposed as a solution to the molecular design problem. However, it has proved challenging to control the design process or incorporate prior knowledge, limiting their practical use in drug discovery. In particular, generative methods have made limited use of three-dimensional (3D) structural information even though this is critical to binding. This work describes a method to incorporate such information and demonstrates the benefit of doing so. We combine an existing graph-based deep generative model, DeLinker, with a convolutional neural network to utilise physically-meaningful 3D representations of molecules and target pharmacophores. We apply our model, DEVELOP, to both linker and R-group design, demonstrating its suitability for both hit-to-lead and lead optimisation. The 3D pharmacophoric information results in improved generation and allows greater control of the design process. In multiple large-scale evaluations, we show that including 3D pharmacophoric constraints results in substantial improvements in the quality of generated molecules. On a challenging test set derived from PDBbind, our model improves the proportion of generated molecules with high 3D similarity to the original molecule by over 300%. In addition, DEVELOP recovers 10 × more of the original molecules compared to the base-line DeLinker method. Our approach is general-purpose, readily modifiable to alternate 3D representations, and can be incorporated into other generative frameworks. Code is available at https://github.com/oxpig/DEVELOP .

Deep Evolutionary Learning for Molecular Design

Evolutionary design of molecules based on deep learning and a genetic algorithm

Probabilistic Generative Deep Learning for Molecular Design

Advances in Deep Learning-Based 3D Molecular Generative Models

Human-level molecular optimization driven by mol-gene evolution

Active Learning-Assisted Directed Evolution

Augmenting Genetic Algorithms with Machine Learning for Inverse Molecular Design

Molecular De Novo Design through Deep Reinforcement Learning

CELLS: Cost-Effective Evolution in Latent Space for Goal-Directed Molecular Generation.

A Transformer-based Generative Model for De Novo Molecular Design

Deep Learning for Multi-Scale Molecular Modeling

Leveraging Latent Evolutionary Optimization for Targeted Molecule Generation

Latent-based Directed Evolution accelerated by Gradient Ascent for Protein Sequence Design

Molecular Design in Synthetically Accessible Chemical Space via Deep Reinforcement Learning

Deep Generative Design with 3D Pharmacophoric Constraints

JANUS: Parallel Tempered Genetic Algorithm Guided by Deep Neural Networks for Inverse Molecular Design

MOLER: Incorporate Molecule-Level Reward to Enhance Deep Generative Model for Molecule Optimization

Evaluation of Reinforcement Learning in Transformer-based Molecular Design

SDEGen: learning to evolve molecular conformations from thermodynamic noise for conformation generation

Machine learning-aided generative molecular design

LEADD: Lamarckian evolutionary algorithm for de novo drug design