3D Molecule Generation From Rigid Motifs via SE(3) Flows

MCML Authors

Marcel Kollovieh

→ Group Stephan Günnemann
 Data Analytics & Machine Learning

Eike Eberhard

→ Group Stephan Günnemann
 Data Analytics & Machine Learning

Stephan Günnemann

Prof. Dr.

Core PI

Data Analytics & Machine Learning

Abstract

Three-dimensional molecular structure generation is typically performed at the level of individual atoms, yet molecular graph generation techniques often consider fragments as their structural units. Building on the advances in frame-based protein structure generation, we extend these fragmentation ideas to 3D, treating general molecules as sets of rigid-body motifs. Utilising this representation, we employ SE(3)-equivariant generative modelling for de novo 3D molecule generation from rigid motifs. In our evaluations, we observe comparable or superior results to state-of-the-art across benchmarks, surpassing it in atom stability on GEOM-DRUGS, while yielding a 2× to 10× reduction in generation steps and offering 3.5× compression in molecular representations compared to the standard atom-based methods.

inproceedings PKE+26

---

GenBio @ICML 2026

Workshop on Generative and Agentic AI for Biology at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available.

---

Authors

R. Poletukhin • M. Kollovieh • E. Eberhard • S. Günnemann

---

Links

URL

---

Research Area

 A3 | Computational Models