揭示了共价对接的极限,并通过共价意识的多任务学习推进了亲和力预测
Jiayang Leng1,2, Zhixuan Huang1,2, Lei Zheng3,4
1Department of Biomedical Engineering, Southern University of Science and Technology, Shenzhen, Guangdong, China.
Physical chemistry chemical physics : PCCP
|February 4, 2026
概括
针对向共价抑制剂 (TCI) 的计算工具尚未得到开发. 这项研究引入了一个新的框架,CovMTL-DTA,该框架显著改善了药物向亲和力预测,并为TCI确定了优先级.
科学领域:
- 药用化学 医学化学
- 计算化学的计算化学
- 药物发现 药物发现 药物发现
背景情况:
- 向共价抑制剂 (TCI) 在药物发现中至关重要,但对它们的结合姿势和亲和力的计算预测仍然具有挑战性.
- 现有的计算工具往往缺乏对共价相互作用的特异性,阻碍了准确的药物向亲和力 (DTA) 预测.
研究的目的:
- 用一个大,精心策划的基准来系统地评估共价对接工具.
- 开发和验证一种新的共价意识的DTA预测框架,以提高TCI的亲和度排名和击中优先级.
主要方法:
- 策划了来自CovalentInDB 2.0的2172个共价蛋白质-连接体复合物的基准,并评估了四个对接引擎.
- 评估了17个共价标的对接分数和实验pIC50值之间的相关性.
- 开发了CovMTL-DTA,这是一个集联联体图,蛋白质嵌入和交叉模式注意力用于共价DTA预测的多任务深度学习模型.
主要成果:
- 博尔茨-2在基准上表现出最佳的姿势复制性能,尽管指出了潜在的数据泄露.
- 来自对接的得分-亲和关系相关性通常很弱,并且依赖于目标 (能否 < 0.2).
- 在一个独立的测试组中,CovMTL-DTA在独立的测试组中实现了~0.77的皮尔森相关性,超过了现有方法.
- 在一个虚拟查活动中,CovMTL-DTA成功地优先考虑了已知的EGFR共价抑制剂.
结论:
- 当前的共价对接工具在姿势预测和亲和度排名方面存在局限性.
- 在预测TCI的药物向亲和力方面,CovMTL-DTA提供了显著的进展.
- 开发的框架改善了命中优先级,并可以加速共价药物发现工作.
相关概念视频
01:29Covalent Bonds
162.9K
Overview
162.9K
01:08Covalent Bonds
11.3K
Overview
When two atoms share electrons to complete their valence shells, they create a covalent bond. An atom's electronegativity—the force with which shared electrons are pulled towards an atom—determines how the electrons are shared. Molecules formed with covalent bonds can be either polar or nonpolar. Atoms with similar electronegativities form nonpolar covalent bonds; the electrons are shared equally. Atoms with different electronegativities share electrons unequally,...
When two atoms share electrons to complete their valence shells, they create a covalent bond. An atom's electronegativity—the force with which shared electrons are pulled towards an atom—determines how the electrons are shared. Molecules formed with covalent bonds can be either polar or nonpolar. Atoms with similar electronegativities form nonpolar covalent bonds; the electrons are shared equally. Atoms with different electronegativities share electrons unequally,...
11.3K
02:18Network Covalent Solids
16.2K
Network covalent solids contain a three-dimensional network of covalently bonded atoms as found in the crystal structures of nonmetals like diamond, graphite, silicon, and some covalent compounds, such as silicon dioxide (sand) and silicon carbide (carborundum, the abrasive on sandpaper). Many minerals have networks of covalent bonds.
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
To break or to melt a covalent network solid, covalent bonds must be broken. Because covalent bonds are relatively strong, covalent network solids are typically...
16.2K
02:46Covalent Bonding and Lewis Structures
61.4K
Compared to ionic bonds, which results from the transfer of electrons between metallic and nonmetallic atoms, covalent bonds result from the mutual attraction of atoms for a “shared” pair of electrons.
61.4K
02:04Covalently Linked Protein Regulators
9.6K
Proteins can undergo many types of post-translational modifications, often in response to changes in their environment. These modifications play an important role in the function and stability of these proteins. Covalently linked molecules include functional groups, such as methyl, acetyl, and phosphate groups, and also small proteins, such as ubiquitin. There are around 200 different types of covalent regulators that have been identified.
These groups modify specific amino acids in a protein....
These groups modify specific amino acids in a protein....
9.6K
02:04Covalently Linked Protein Regulators
2.0K
2.0K