Comparing deep learning models for designing new drug molecules
A Systematic Evaluation of Molecule Generation Models for De Novo Drug Design: From Benchmarks to Practical Insights
Machine Learning
Summary
Finding new medicines involves creating new molecules that can work effectively inside the body. The authors looked at 82 different computer models that use artificial intelligence to design these molecules from scratch. They compared how these models work, what tools and data they use, and where they perform well or poorly. Their study helps people understand the strengths and trade-offs of different AI methods for drug discovery and points out future steps to improve how these models work with real 3D molecular data and biological targets.
What this means in practice
- •For pharmaceutical ai developers: Improve molecular design pipelines by selecting the best performing generative models for specific drug discovery goals based on a comprehensive model comparison.
- •For chemical informatics teams: Use consolidated benchmark data and evaluation metrics to standardize molecule generation practices and enhance reproducibility across drug design projects.
A survey. It maps existing work.
Authors
Xinrui Xu, Xueer Wang, Dan Luo, Sisi Yuan, Xuan Lin
Abstract
Molecule generation has emerged as a powerful computational tool for de novo drug design, enabling the exploration of chemical space beyond the limits of conventional virtual screening. The field has progressed rapidly, driven by advances in molecular representations, generative architectures, and target-aware modeling strategies. However, existing reviews typically address specific model families or application scenarios in isolation, rather than offering an integrated perspective on how these components collectively form a coherent generation workflow. In this review, we present a comprehensive evaluation of molecule generation models for de novo drug design, covering 82 methods across five deep generative frameworks, including recurrent neural network (RNN)- and Transformer-based models, variational autoencoders (VAEs), generative adversarial networks (GANs), flow-based models, and diffusion models. We first summarize widely used benchmarks and molecular representations, and then examine the methodological principles underlying both general and pocket-conditioned generation. A central contribution of this work is a systematic synthesis and comparative analysis of reported performance across commonly used benchmarks and evaluation metrics. We also summarize representative experimentally validated case studies. Looking ahead, we discuss future directions in standardized 3D data, interaction-aware generation, receptor flexibility, and multi-objective molecular design, with the aim of improving the reliability and experimental relevance of molecule generation. All collected benchmark resources, evaluation metrics, and model references are provided in a publicly accessible repository at https://github.com/JacklinGroup/molecule-generation-review.