杜里安:基于结构的3D分子生成的全面基准
Dou Nie1, Huifeng Zhao1, Odin Zhang1
1Innovation Institute for Artificial Intelligence in Medicine of Zhejiang University, College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, 310058 Zhejiang, China.
Journal of chemical information and modeling
|December 16, 2024
概括
杜兰,一个新的框架,使用多种指标评估基于结构的3D分子生成模型. 它揭示了药物发现中新性和实用性的平衡的局限性,强调了多目标优化.
科学领域:
- 计算化学计算化学
- 药物发现 药物发现 药物发现
- 化学中的人工智能.
背景情况:
- 深度神经网络用于3D分子生成,利用目标结构信息.
- 这些模型的当前评估方法受到数据偏差和单个指标的限制.
- 由于评估标准不一致,比较生成模型具有挑战性.
研究的目的:
- 介绍Durian,一个基于结构的3D分子生成的全面评估框架.
- 使用多度指标方法评估六种领先的3D分子生成方法.
- 为药物设计生成模型的选择和优化提供指导.
主要方法:
- 开发了Durian,结合了蛋白质-配体数据,实验亲和力,以及物理化学/几何指标.
- 在"Dock"和"Score"模式中使用了三种对接方法 (QuickVina2,Surflex,Gnina) 来进行绑定亲和度评估.
- 应用Durian对六个3D分子生成模型进行了基准测试:LiGAN,Pocket2Mol,DiffSBDD,SBDD,GraphBP和SurfGen.
主要成果:
- 大多数模型都产生了具有合理性质的药物样分子,但在新性,理性和合成可访问性方面遇到了困难.
- 杜里安揭示了各种模型的不同局限性,强调了需要多目标优化.
- SurfGen和SBDD的整体表现强,但需要在构造合理性方面进行改进.
结论:
- 杜里安为评估3D分子生成模型提供了一个强大的框架.
- 多目标优化对于开发用于药物发现的实用3D生成模型至关重要.
- 该框架为药物设计管道中选择和完善生成模型提供了指导.
相关概念视频
02:00Molecular Models
37.9K
Physical models representing molecular architectures of chemical compounds play essential roles in understanding chemistry. The use of molecular models makes it easier to visualize the structures and shapes of atoms and molecules.
37.9K
02:06Newman Projections
16.4K
Different notations are used to represent the three-dimensional structure of molecules on two-dimensional surfaces. One of the most commonly used representations is the dash-wedge formula. The dashed wedges, solid wedges, and the plane lines indicate the groups situated behind the plane, coming out of the plane, and in the plane, respectively.
The organic molecules rotate across the single bonds leading to numerous temporary three-dimensional structures of varying energy known as...
The organic molecules rotate across the single bonds leading to numerous temporary three-dimensional structures of varying energy known as...
16.4K