使用深度变异异常生成的Fuzz测试分子表示
Victor H R Nogueira1,2, Rishabh Sharma3, Rafael V C Guido1
1São Carlos Institute of Physics, University of São Paulo, São Paulo 13563-120, Brazil.
Journal of chemical information and modeling
|February 5, 2025
概括
研究人员使用深度学习模型在SELFIES格式中找到无效的分子字符串,揭示了分子表示的弱点,并建议改进以提高可靠性.
科学领域:
- 计算化学是一种计算化学.
- 机器学习是机器学习.
- 化学信息学 化学信息学
背景情况:
- 分子表示对于化学信息学和药物发现至关重要.
- 现有的分子字符串格式需要严格的验证方法.
- 深度学习模型为测试数据完整性提供了新的途径.
研究的目的:
- 为了应力测试自我引用嵌入字符串 (SELFIES) 分子格式的稳定性.
- 识别和分析不正常的SELFIES字符串,这些字符串不会转换为有效的化学结构.
- 探索无监督深度学习对验证分子表示的实用性.
主要方法:
- 使用一个变化自动编码器 (VAE),一个无监督的深度学习模型,生成异常的SELFIES字符串.
- 分析了VAE的潜空间,以确定生产不可转换的SELFIES的区域.
- 调查了生成的SELFIES字符串无效的结构和化学原因.
主要成果:
- 成功生成了无法转换为有效的SMILES字符串的异常SELFIES字符串.
- 确定了与生成无效SELFIES相关的VAE潜伏空间中的特定区域.
- 发现并非所有SELFIES字符串都可以转换为有效的SMILES,这挑战了以前的假设.
结论:
- 变量自动编码器是实证压力测试分子表示强度的有效工具.
- 该研究强调了对SELFIES等分子字符串格式的改进验证协议的需求.
- 结果提供了关于SELFIES字符串无效性的见解,并建议修改以提高可靠性.
相关概念视频
Molecular Models
37.8K
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.
37.8K
Predicting Molecular Geometry
34.0K
VSEPR Theory for Determination of Electron Pair Geometries
34.0K
Molecular Geometry and Dipole Moments
12.5K
The VSEPR theory can be used to determine the electron pair geometries and molecular structures as follows:
12.5K
Hybridization of Atomic Orbitals II
31.7K
sp3d and sp3d 2 Hybridization
31.7K
Valence Bond Theory and Hybridized Orbitals
18.7K
According to valence bond theory, a covalent bond results when: (1) an orbital on one atom overlaps an orbital on a second atom, and (2) the single electrons in each orbital combine to form an electron pair. The strength of a covalent bond depends on the extent of overlap of the orbitals involved. Maximum overlap is possible when the orbitals overlap on a direct line between the two nuclei.
A σ bond (single bond in a Lewis structure) is a covalent bond in which the electron density is...
A σ bond (single bond in a Lewis structure) is a covalent bond in which the electron density is...
18.7K
Newman Projections
16.3K
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...
16.3K


