阿尔法预测是有价值的假设,并加速,但不能取代实验结构确定
Thomas C Terwilliger1,2, Dorothee Liebschner3, Tristan I Croll4
1New Mexico Consortium, Los Alamos, NM, USA. tterwilliger@newmexicoconsortium.org.
Nature methods
|November 30, 2023
概括
像AlphaFold这样的人工智能 (AI) 蛋白质结构预测是强大的工具,但其准确性各不相同. 实验验证对于细节至关重要,特别是那些没有被人工智能模型考虑的细节.
科学领域:
- 结构生物学是结构生物学.
- 计算生物学是一种计算生物学.
- 生物物理学的生物物理.
背景情况:
- 人工智能驱动的蛋白质结构预测,以AlphaFold为例,已经改变了结构生物学.
- 然而,这些预测的准确性并不统一,并且不考虑连接物,修饰或环境影响.
研究的目的:
- 通过将它们与实验晶体学数据进行比较来评估AlphaFold预测的可靠性.
- 了解AI在蛋白质结构预测中的局限性.
主要方法:
- 将AlphaFold生成的蛋白质结构预测与实验晶体图的比较.
- 在全球 (域面方向) 和地方 (构造) 尺度上分析差异.
主要成果:
- 在许多情况下,AlphaFold的预测与实验地图密切匹配.
- 在某些情况下观察到显著的偏差,影响全球和本地结构特征,即使对于高可信度预测.
结论:
- 应将AlphaFold预测视为需要实验验证的有价值的假设.
- 预测信心水平对于解释至关重要;实验结构的确定对于验证细节至关重要,特别是涉及未建模的相互作用的细节.
相关概念视频
01:49Conserved Binding Sites
4.2K
Many proteins’ biological role depends on their interactions with their ligands, small molecules that bind to specific locations on the protein known as ligand-binding sites. Ligand-binding sites are often conserved among homologous proteins as these sites are critical for protein function.
Binding sites are often located in large pockets, and if their location on a protein’s surface is unknown, it can be predicted using various approaches. The energetic method computationally...
Binding sites are often located in large pockets, and if their location on a protein’s surface is unknown, it can be predicted using various approaches. The energetic method computationally...
4.2K
01:22Protein Folding
118.2K
Overview
118.2K
01:13Protein Organization
138.0K
Overview
138.0K