Reproduce “3D Molecule Generation from Rigid Motifs via SE(3) Flows” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Roman Poletukhin, Marcel Kollovieh, Eike Eberhard, Stephan Günnemann
MOTIFLOW generates 3D molecules by selecting reusable rigid chemical motifs and jointly predicting each motif’s identity and pose, then reconstructing atoms from stored canonical geometry. On GEOM-DRUGS, at 100 sampling steps, it reports 95.0 ± 0.0% atom stability and 81.2 ± 0.3% valid-and-connected samples, versus END’s 87.2 ± 0.1% and 73.7 ± 0.4%. This comparison uses bonds inferred from distances.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.