# Diffusion-based Evolutionary Optimization for 3D Multi-Objective Molecular Generation

arXiv: https://arxiv.org/abs/2505.11037 · HTML: https://mlsanity.com/paper/2505.11037/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Flow matching, time series and science, one of 14 directions in Diffusion models.
- Builds on 14 of its 35 references that are in the map.
- First posted on arXiv 16 May 2025.

## Placed in

- Diffusion models / [Flow matching, time series and science](https://mlsanity.com/diffusion/flow-matching-time-series-and-science/)

## Builds on (most cited first, 14 of 14)

- [Score-Based Generative Modeling through Stochastic Differential Equations](https://mlsanity.com/paper/2011.13456/) (2020)
- [RePaint: Inpainting using Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2201.09865/) (2022)
- [Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds](https://mlsanity.com/paper/1802.08219/) (2018)
- [Equivariant Diffusion for Molecule Generation in 3D](https://mlsanity.com/paper/2203.17003/) (2022)
- [GEOM: Energy-annotated molecular conformations for property prediction and molecular generation](https://mlsanity.com/paper/2006.05531/) (2020)
- [Optimization of Molecules via Deep Reinforcement Learning](https://mlsanity.com/paper/1810.08678/) (2018)
- [Geometric Latent Diffusion Models for 3D Molecule Generation](https://mlsanity.com/paper/2305.01140/) (2023)
- [TFG: Unified Training-Free Guidance for Diffusion Models](https://mlsanity.com/paper/2409.15761/) (2024)
- [Equivariant Energy-Guided SDE for Inverse Molecular Design](https://mlsanity.com/paper/2209.15408/) (2022)
- [Efficient Evolutionary Search Over Chemical Space with Large Language Models](https://mlsanity.com/paper/2406.16976/) (2024)
- [Geometry-Complete Diffusion for 3D Molecule Generation and Optimization](https://mlsanity.com/paper/2302.04313/) (2023)
- [Computer-Aided Multi-Objective Optimization in Small Molecule Discovery](https://mlsanity.com/paper/2210.07209/) (2022)
- [Text-Guided Molecule Generation with Diffusion Language Model](https://mlsanity.com/paper/2402.13040/) (2024)
- [Predicting Side Effect of Drug Molecules using Recurrent Neural Networks](https://mlsanity.com/paper/2305.10473/) (2023)

## Built on by (most cited first, 2 of 2)

- [Diffusion Models at the Drug Discovery Frontier: A Review on Generating Small Molecules versus Therapeutic Peptides](https://mlsanity.com/paper/2511.00209/) (2025)
- [DiffGA: Steering Frozen Diffusion Models for Materials Design via Evolutionary Search](https://mlsanity.com/paper/2610.14999/) (2026)

## Abstract (arXiv)

Optimizing conflicting molecular properties while strictly adhering to complex 3D structural constraints constitutes a challenging Constrained Multi-Objective Optimization Problem (CMOP). Traditional Evolutionary Algorithms (EAs) destroy chemical valency in 3D space, whereas 3D diffusion models act as rigid generators requiring costly retraining for novel objectives. To bridge this gap, we propose a progressive algorithmic suite. First, we introduce the Evolutionary-Guided Diffusion (EGD) operator, which executes crossover and mutation at an optimally calibrated noise level, leveraging a pre-trained denoising network to project chimeric states back onto the valid chemical manifold. Second, to combat the severe loss of molecular structural diversity inherent in traditional EMO frameworks, we design a Structure-Aware Environmental Selection (SAES) mechanism that explicitly enforces structural distinctiveness. Finally, synergizing EGD and SAES, we develop the Diffusion-based Evolutionary Molecular Optimization (DEMO) framework for CMOPs. To safely navigate disjoint feasible regions, DEMO employs a tri-population architecture with distinct goals: exploring novel chemical scaffolds, refining partially assembled intermediates, and fine-tuning perfectly feasible elite molecules. Extensive experiments across single-property targeting, unconstrained MOPs, multi-fragment CMOPs, and 3D protein-ligand docking demonstrate that our method comprehensively outperforms state-of-the-art baselines and traditional EMO frameworks. Operating entirely zero-shot, this suite consistently discovers highly diverse, chemically valid Pareto frontiers.