AI framework streamlines drug discovery with synthesizable molecule design

Researchers have developed a new generative design framework that addresses the challenge of creating molecules that are both optimal in properties and easily synthesizable, a key hurdle in drug discovery. This framework allows for steerable and granular control over the synthesizability of generated molecules, enabling the incorporation of specific reaction constraints and building blocks. The system was successfully applied in an in-house campaign targeting BRD4, leading to the design, synthesis, and identification of two micromolar binders. Furthermore, the approach demonstrated efficiency in navigating ultra-large chemical spaces, identifying a micromolar Wee1 binder from a library of 142 billion molecules using minimal computational resources. AI

影响 Enables faster and more efficient identification of novel, synthesizable drug candidates by overcoming key bottlenecks in molecular design and screening.

排序理由 Academic paper detailing a new AI framework for molecular design. [lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

• BRD4
• Chemspace
• Freedom 4.0
• WEE1

• 论文
• 产品

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →