pith. sign in

arxiv: 2209.15408 · v3 · pith:D6RDRZSTnew · submitted 2022-09-30 · ⚛️ physics.chem-ph · cs.LG· q-bio.BM

Equivariant Energy-Guided SDE for Inverse Molecular Design

classification ⚛️ physics.chem-ph cs.LGq-bio.BM
keywords moleculareegsdeenergydesigninversepropertiesenergy-guidedequivariant
0
0 comments X
read the original abstract

Inverse molecular design is critical in material science and drug discovery, where the generated molecules should satisfy certain desirable properties. In this paper, we propose equivariant energy-guided stochastic differential equations (EEGSDE), a flexible framework for controllable 3D molecule generation under the guidance of an energy function in diffusion models. Formally, we show that EEGSDE naturally exploits the geometric symmetry in 3D molecular conformation, as long as the energy function is invariant to orthogonal transformations. Empirically, under the guidance of designed energy functions, EEGSDE significantly improves the baseline on QM9, in inverse molecular design targeted to quantum properties and molecular structures. Furthermore, EEGSDE is able to generate molecules with multiple target properties by combining the corresponding energy functions linearly.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. GeoCycler: Reward-Aligned 3D Diffusion for Constraint-Conditioned Cyclic Peptide Design

    cs.CE 2026-05 unverdicted novelty 7.0

    GeoCycler aligns latent diffusion models via reward-weighted training with a type-gated stair reward to raise cyclic peptide closure rates across multiple topologies on the LNR benchmark.

  2. APCyc: Property-Informed Design of Cyclic Peptides via Automated Cyclization

    cs.AI 2026-06 unverdicted novelty 6.0

    APCyc is a target-aware generative model for de novo cyclic peptide design that adds cyclization-site encoding and Bayesian guidance to jointly optimize physicochemical properties.

  3. ParetoPilot: Zero-Surrogate Offline Multi-Objective Optimization via Infer-Perturb-Guide Diffusion

    cs.LG 2026-06 unverdicted novelty 6.0

    ParetoPilot is a zero-surrogate diffusion framework for offline MOO that uses an IPG engine to steer generation via inferred objective directions and orthogonal perturbations, outperforming 14 baselines on 51 tasks.

  4. Latent Diffusion Pretraining for Crystal Property Prediction

    cs.LG 2026-05 unverdicted novelty 6.0

    CrysLDNet combines VAE and latent diffusion pretraining on unlabeled crystals to improve graph encoder performance on property prediction by about 4-5% on JARVIS and MP datasets.