ClickGen: Directed exploration of synthesizable chemical space via modular reactions and reinforcement learning DOI Creative Commons

Mingyang Wang,

Shuai Li, Jike Wang

et al.

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Nov. 22, 2024

Despite the significant potential of generative models, low synthesizability many generated molecules limits their real-world applications. In response to this issue, we develop ClickGen, a deep learning model that utilizes modular reactions like click chemistry assemble and incorporates reinforcement along with inpainting technique ensure proposed display high diversity, novelty strong binding tendency. ClickGen demonstrates superior performance over other reaction-based models in terms novelty, synthesizability, docking conformation similarity for existing binders targeting three proteins. We then proceeded conduct wet-lab validation on ClickGen's poly adenosine diphosphate-ribose polymerase 1. Due guaranteed model-generated synthetic routes reference, successfully produced tested bioactivity these novel compounds just 20 days, much faster than typically expected time frame when handling sufficiently molecules. assays, two lead demonstrated anti-proliferative efficacy against cancer cell lines, toxicity, nanomolar-level inhibitory activity PARP1. demonstrate related may represent new paradigm molecular generation, bringing AI-driven, automated experimentation closed-loop design closer realization. Generative face challenges Here, authors using generate highly diverse, novel, synthesizable

Language: Английский

Engaging an engineered PARP-2 catalytic domain mutant to solve the complex structures harboring approved drugs for structure analyses DOI
Xiaoyu Wang, Jie Zhou, Bailing Xu

et al.

Bioorganic Chemistry, Journal Year: 2025, Volume and Issue: unknown, P. 108471 - 108471

Published: April 1, 2025

Language: Английский

Citations

0
A Prostate-Specific Membrane Antigen-Targeting Small Molecule–Drug Conjugate Strategy to Overcome the Hematological Toxicity of Olaparib DOI
Qi Chen, Zhenying Wu,

Haiying Zhu

et al.

Journal of Medicinal Chemistry, Journal Year: 2024, Volume and Issue: 67(21), P. 19586 - 19611

Published: Nov. 1, 2024

PARP inhibitors have gained attention in the treatment of metastatic castration-resistant prostate cancer, but approximately half patients to abort due severe hematological toxicity. Herein, we proposed a prostate-specific membrane antigen (PSMA)-targeting small molecule-drug conjugate (SMDC) strategy address this issue. This led

Language: Английский

Citations

2
The Potential of PARP Inhibitors as Antitumor Drugs and the Perspective of Molecular Design DOI
Yinghan Wang, Jingtao Zhang, Xiaochen Wu

et al.

Journal of Medicinal Chemistry, Journal Year: 2024, Volume and Issue: 68(1), P. 18 - 48

Published: Dec. 26, 2024

PARP (poly-ADP ribose polymerase) has received widespread attention in cancer treatment. Research shown that plays a crucial role DNA damage repair and become popular target for drug design. Based on the mechanism of "synthetic lethality", multiple PARPis (PARP inhibitors) have been launched treatment BRCA deficient tumors. For example, approved significant potential treatment, particularly breast cancers associated with BRCA1/BRCA2 deficiencies. However, clinical efficacy safety inhibitors different remain issues cannot be overlooked. The design aims to eliminate their resistance broaden application scope. Designing selective PARP-1 is also strategy. PROTACs (Proteolysis Targeting Chimeras) degrade novel

Language: Английский

Citations

1
Discovery of Pyrazolo[1,5,4-de]quinoxalin-2(3H)-one Derivatives as Highly Potent and Selective PARP1 Inhibitors DOI

Shanyun Gao,

Yingjie Hou,

Yanxiao Xu

et al.

Journal of Medicinal Chemistry, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 21, 2024

Poly-ADP-ribose-polymerase 1/2 (PARP1/2) inhibitors have been approved for cancers with homologous recombination deficiency (HRD). However, their narrow therapeutic indexes largely due to hematologic toxicities limited clinical usefulness. Developing selective PARP1 has emerged as an attractive strategy achieve equivalent antitumor activity while alleviating the hematological toxicity caused by PARP2 inhibition. Herein, we report discovery of pyrazolo[1,5,4-de]quinoxalin-2(3H)-one 30 a novel inhibitor. formed tighter PARP1-DNA trapping than AZD9574, leading better potency in inhibiting cancer cell proliferation. achieved tumor regression BRCA1-mutated MDA-MB-436 xenograft model and showed synergistic efficacy combination carboplatin SUM149PT model. In rat study, exhibited minimal impact on parameters at 25 mg/kg, AZD5305 1 mg/kg 56.5% reduction reticulocyte. Taken together, discovered compound index superior that AZD9574 preclinical setting.

Language: Английский

Citations

0
Catalyst-Free Construction of Imidazole-Pyrrolo[1,2-a]pyrazine Hybrid, 2,6-Disubstituted Imidazo[1,2-a]pyrrolo[2,1-c]pyrazine via Regioselective Annulative Functionalizations DOI

Hyunjin Oh,

Daniel C. Ku,

M. JUNG

et al.

The Journal of Organic Chemistry, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 26, 2024

Highly efficient catalyst-free annulative functionalization approaches to a novel imidazole-pyrrolo[1,2-

Language: Английский

Citations

0
ClickGen: Directed exploration of synthesizable chemical space via modular reactions and reinforcement learning DOI Creative Commons

Mingyang Wang,

Shuai Li, Jike Wang

et al.

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Nov. 22, 2024

Despite the significant potential of generative models, low synthesizability many generated molecules limits their real-world applications. In response to this issue, we develop ClickGen, a deep learning model that utilizes modular reactions like click chemistry assemble and incorporates reinforcement along with inpainting technique ensure proposed display high diversity, novelty strong binding tendency. ClickGen demonstrates superior performance over other reaction-based models in terms novelty, synthesizability, docking conformation similarity for existing binders targeting three proteins. We then proceeded conduct wet-lab validation on ClickGen's poly adenosine diphosphate-ribose polymerase 1. Due guaranteed model-generated synthetic routes reference, successfully produced tested bioactivity these novel compounds just 20 days, much faster than typically expected time frame when handling sufficiently molecules. assays, two lead demonstrated anti-proliferative efficacy against cancer cell lines, toxicity, nanomolar-level inhibitory activity PARP1. demonstrate related may represent new paradigm molecular generation, bringing AI-driven, automated experimentation closed-loop design closer realization. Generative face challenges Here, authors using generate highly diverse, novel, synthesizable

Language: Английский

Citations

0