来自不同来源的IC50或K值的组合是造成大量噪音的来源
Gregory A Landrum1, Sereina Riniker1
1Department of Chemistry and Applied Biosciences, ETH Zurich, Vladimir-Prelog-Weg 2, 8093 Zurich, Switzerland.
Journal of chemical information and modeling
|February 23, 2024
概括
结合来自不同测试的生物活性数据,如IC50和Ki,引入了显著的噪音,损害了机器学习模型的验证. 仔细的数据策划,特别是使用最大策划策略,对于可靠的生物活性预测至关重要.
科学领域:
- 计算化学是一种计算化学.
- 化学信息学 化学信息学
- 生物活性的预测预测.
背景情况:
- 用于生物活性预测的机器学习 (ML) 模型需要大量的,经过验证的数据集.
- 研究人员经常将来自各种文献分析的IC50和Ki数据结合起来,包括ChEMBL,用于ML模型培训和验证.
- 结合来自不同试验的数据,即使是针对相同的目标,由于潜在的不兼容性,也会带来重大科学风险.
研究的目的:
- 通过结合来自不同试验的生物活性数据来量化引入的噪音.
- 评估数据策略对 ML 应用组合数据集质量的影响.
- 提出和验证"最大限度的策划"方法来提高数据质量.
主要方法:
- 对相同的目标和化合物在ChEMBL数据库中的重叠测定分析.
- 通过比较不同测试中的IC50和Ki测量来估计数据噪声.
- 在最小策划和最大策划设置之间比较数据质量.
- 使用肯德尔的陶氏相关系数进行统计分析.
主要成果:
- 对IC50试验的最小化导致差异很小:65%的点差>0.3日志单位,27%的点差>1日志单位,肯德尔的tau为0.51.
- 最大策划 (匹配测试元数据) 改进一致性:48%的区别为>0.3日志单位,13%为>1日志单位,肯德尔的tau为0.71,尽管数据集较小.
- 结合不同的文献Ki测试时,观察到类似的噪声水平.
- 该研究强调了生物活性数据集组合中的大量噪音.
结论:
- 结合不同试验的生物活性数据而没有仔细的固化,引入了显著的噪音,影响了ML模型的可靠性.
- 一个最大度的策划策略,需要匹配的测试元数据,大大提高了数据质量,但减少了数据集大小.
- 研究人员必须在构建和验证用于生物活性预测的ML模型时实施严格的数据策划实践.
- 提供开源代码和ChEMBL查询以促进最大限度的策划方法.
相关概念视频
Inductively Coupled Plasma-Mass Spectrometry (ICP-MS): Interferences
469
Inductively coupled plasma–mass spectrometry (ICP–MS) is a highly selective and sensitive technique for accurate elemental analysis. Though the analysis of ICP–MS mass spectra is comparatively straightforward, it is affected by spectroscopic and non-spectroscopic interferences. Spectroscopic interferences arise when the plasma contains ionic species with an m/z value the same as the analyte ion. Spectroscopic interference can be categorized as isobaric, polyatomic ions, and...
469
¹³C NMR: ¹H–¹³C Decoupling
1.1K
The probability of having two carbon-13 atoms next to each other is negligible because of the low natural abundance of carbon-13. Consequently, peak splitting due to carbon-carbon spin-spin coupling is not observed in spectra. However, protons up to three sigma bonds away split the carbon signal according to the n+1 rule, resulting in complicated spectra.
A broadband decoupling technique is used to simplify these complex, sometimes overlapping, signals. Broadband decoupling relies on a...
A broadband decoupling technique is used to simplify these complex, sometimes overlapping, signals. Broadband decoupling relies on a...
1.1K
Superposition Theorem
598
The superposition principle is a fundamental concept stating that in a linear circuit, the voltage across (or current through) an element can be determined by summing the individual contributions of each independent source acting in isolation. When dealing with linear circuits containing multiple independent sources, this principle serves as a valuable tool for analysis. To apply the superposition principle effectively, one should focus on a single independent source at a time while...
598
Superposition Theorem for AC Circuits
667
Consider encountering a circuit in a steady state where all its inputs are sinusoidal, yet they do not all possess the same frequency. Such a circuit is not classified as an alternating current (AC) circuit, and consequently, its currents and voltages will not exhibit sinusoidal behavior. However, this circuit can be analyzed using the principle of superposition.
The principle of superposition stipulates that the output of a linear circuit with several concurrent inputs is equivalent to the...
The principle of superposition stipulates that the output of a linear circuit with several concurrent inputs is equivalent to the...
667


