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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Molecular conformation generation aims to generate three-dimensionalcoordinates of all the atoms in a molecule and is an important task inbioinformatics and pharmacology. Previous methods usually first predict theinteratomic distances, the gradients of interatomic distances or the localstructures (e.g., torsion angles) of a molecule, and then reconstruct its 3Dconformation. How to directly generate the conformation without the aboveintermediate values is not fully explored. In this work, we propose a methodthat directly predicts the coordinates of atoms: (1) the loss function isinvariant to roto-translation of coordinates and permutation of symmetricatoms; (2) the newly proposed model adaptively aggregates the bond and atominformation and iteratively refines the coordinates of the generatedconformation. Our method achieves the best results on GEOM-QM9 and GEOM-Drugsdatasets. Further analysis shows that our generated conformations have closerproperties (e.g., HOMO-LUMO gap) with the groundtruth conformations. Inaddition, our method improves molecular docking by providing better initialconformations. All the results demonstrate the effectiveness of our method andthe great potential of the direct approach. The code is released athttps://github.com/DirectMolecularConfGen/DMCG

Direct Molecular Conformation Generation