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

In the past decade, Artificial Intelligence driven drug design and discoveryhas been a hot research topic, where an important branch is molecule generationby generative models, from GAN-based models and VAE-based models to the latestdiffusion-based models. However, most existing models pursue only the basicproperties like validity and uniqueness of the generated molecules, a few gofurther to explicitly optimize one single important molecular property (e.g.QED or PlogP), which makes most generated molecules little usefulness inpractice. In this paper, we present a novel approach to generating moleculeswith desirable properties, which expands the diffusion model framework withmultiple innovative designs. The novelty is two-fold. On the one hand,considering that the structures of molecules are complex and diverse, andmolecular properties are usually determined by some substructures (e.g.pharmacophores), we propose to perform diffusion on two structural levels:molecules and molecular fragments respectively, with which a mixed Gaussiandistribution is obtained for the reverse diffusion process. To get desirablemolecular fragments, we develop a novel electronic effect based fragmentationmethod. On the other hand, we introduce two ways to explicitly optimizemultiple molecular properties under the diffusion model framework. First, aspotential drug molecules must be chemically valid, we optimize molecularvalidity by an energy-guidance function. Second, since potential drug moleculesshould be desirable in various properties, we employ a multi-objectivemechanism to optimize multiple molecular properties simultaneously. Extensiveexperiments with two benchmark datasets QM9 and ZINC250k show that themolecules generated by our proposed method have better validity, uniqueness,novelty, Fr\'echet ChemNet Distance (FCD), QED, and PlogP than those generatedby current SOTA models.

Diffusing on Two Levels and Optimizing for Multiple Properties: A Novel Approach to Generating Molecules with Desirable Properties