We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

physics.chem-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Physics > Chemical Physics

Title: Torsional Diffusion for Molecular Conformer Generation

Abstract: Molecular conformer generation is a fundamental task in computational chemistry. Several machine learning approaches have been developed, but none have outperformed state-of-the-art cheminformatics methods. We propose torsional diffusion, a novel diffusion framework that operates on the space of torsion angles via a diffusion process on the hypertorus and an extrinsic-to-intrinsic score model. On a standard benchmark of drug-like molecules, torsional diffusion generates superior conformer ensembles compared to machine learning and cheminformatics methods in terms of both RMSD and chemical properties, and is orders of magnitude faster than previous diffusion-based models. Moreover, our model provides exact likelihoods, which we employ to build the first generalizable Boltzmann generator. Code is available at this https URL
Subjects: Chemical Physics (physics.chem-ph); Machine Learning (cs.LG); Biomolecules (q-bio.BM)
Cite as: arXiv:2206.01729 [physics.chem-ph]
  (or arXiv:2206.01729v1 [physics.chem-ph] for this version)

Submission history

From: Bowen Jing [view email]
[v1] Wed, 1 Jun 2022 04:30:41 GMT (2443kb,D)

Link back to: arXiv, form interface, contact.