English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Designing new molecules is essential for drug discovery and material science.Recently, deep generative models that aim to model molecule distribution havemade promising progress in narrowing down the chemical research space andgenerating high-fidelity molecules. However, current generative models onlyfocus on modeling either 2D bonding graphs or 3D geometries, which are twocomplementary descriptors for molecules. The lack of ability to jointly modelboth limits the improvement of generation quality and further downstreamapplications. In this paper, we propose a new joint 2D and 3D diffusion model(JODO) that generates complete molecules with atom types, formal charges, bondinformation, and 3D coordinates. To capture the correlation between moleculargraphs and geometries in the diffusion process, we develop a Diffusion GraphTransformer to parameterize the data prediction model that recovers theoriginal data from noisy data. The Diffusion Graph Transformer interacts nodeand edge representations based on our relational attention mechanism, whilesimultaneously propagating and updating scalar features and geometric vectors.Our model can also be extended for inverse molecular design targeting single ormultiple quantum properties. In our comprehensive evaluation pipeline forunconditional joint generation, the results of the experiment show that JODOremarkably outperforms the baselines on the QM9 and GEOM-Drugs datasets.Furthermore, our model excels in few-step fast sampling, as well as in inversemolecule design and molecular graph generation. Our code is provided inhttps://github.com/GRAPH-0/JODO.

Learning Joint 2D & 3D Diffusion Models for Complete Molecule Generation