跳到正文
Molence· @MolenceBio · X·· 17 天前AI 评分54
AI 导读

Molence 使用 Mol-1 的早期检查点构建 de novo VHH 设计工作流,在 17 个不同靶点上均获得结合命中。初始靶点级命中率为 6% 至 63%,覆盖可溶性蛋白、受体胞外域和 GPCR,包括 BCMA、4-1BB、TNFR2、IL-4Rα、CCR8、CXCR4、CXCR6 和 CCR5。作者称这是一组经过实验测试的多样化靶点设计。

正文 · 原文

3/15 🧪 Binding hits across all 17 diverse targets.
Using an earlier checkpoint of Mol-1, we built a de novo VHH design workflow. Initial target-level hit rates ranged from 6% to 63%, spanning soluble proteins, receptor extracellular domains and GPCRs—including BCMA, 4-1BB, TNFR2, IL-4Rα, CCR8, CXCR4, CXCR6 and CCR5. This is experimentally tested design across a diverse target panel. #ProteinDesign

引用Molence@MolenceBio
2/15 🧩 A 60,120- molecular assembly, with structure prediction on a single NVIDIA H20. Mol-1-based long running agent outperforms Anthropic’s reported results on 9 of 10 evaluated assemblies. Eight achieve TM-score ≥ 0.90. #AIforScience
在 X 查看被引用的帖子

来源:Molence · x.com