当前位置: X-MOL 学术Int. J. Environ. Res. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
The System of Self-Consistent of Models: A New Approach to Build Up and Validation of Predictive Models of the Octanol/Water Partition Coefficient for Gold Nanoparticles
International Journal of Environmental Research ( IF 2.6 ) Pub Date : 2021-06-05 , DOI: 10.1007/s41742-021-00346-w
Alla P. Toropova , Andrey A. Toropov

The octanol/water partition coefficient for nanoparticles is an important criterion to estimate the ecological fate of these new fast disseminate in everyday life substances. The validation of the predictive potential of models is a key component of quantitative structure–property relationships (QSPRs). The so-called system of self-consistent of the models is, maybe, a new approach for the validation of the predictive potential. The measure of self-consistency is the average of the correlation coefficient on different validation sets observed for different models. The aim of the present study was the checking up the suitability of the self-consistency of models obtained by two approaches to select a better modelling approach of octanol/water partition coefficients for gold nanoparticles (GNPs). The above models are based on a representation of GNPs by so-called quasi-SMILES which are special sequences of symbols which translate data on architecture and conditions of acting of GNPs. The first approach is the Monte Carlo optimization of the correlation coefficient between the observed and predicted endpoint. The second is the modification of the first approach via an additional criterion, so-called the index of the ideality of correlation. The self-consistency and predictive potential of the second approach are better. The system of self-consistency of models is studied for modeling of octanol/water partition coefficient of GNPs, for the first time.

Graphic Abstract



中文翻译:

模型自洽系统:一种建立和验证金纳米颗粒辛醇/水分配系数预测模型的新方法

纳米颗粒的辛醇/水分配系数是评估这些新的快速传播的日常生活物质的生态命运的重要标准。模型预测潜力的验证是定量结构-性质关系 (QSPR) 的关键组成部分。所谓的模型自洽系统可能是一种验证预测潜力的新方法。自洽性的度量是针对不同模型观察到的不同验证集的相关系数的平均值。本研究的目的是检查通过两种方法获得的模型的自洽性的适用性,以选择更好的金纳米粒子 (GNP) 辛醇/水分配系数建模方法。上述模型基于所谓的准微笑对 GNP 的表示,准微笑是特殊的符号序列,用于转换有关 GNP 的架构和作用条件的数据。第一种方法是对观察到的和预测的端点之间的相关系数进行蒙特卡罗优化。第二种是通过附加标准对第一种方法的修改,即所谓的相关理想指数。第二种方法的自洽性和预测潜力更好。首次研究了模型自洽系统对GNPs辛醇/水分配系数的建模。第一种方法是对观察到的和预测的端点之间的相关系数进行蒙特卡罗优化。第二种是通过附加标准对第一种方法的修改,即所谓的相关理想指数。第二种方法的自洽性和预测潜力更好。首次研究了模型自洽系统对GNPs辛醇/水分配系数的建模。第一种方法是对观察到的和预测的端点之间的相关系数进行蒙特卡罗优化。第二种是通过附加标准对第一种方法的修改,即所谓的相关理想指数。第二种方法的自洽性和预测潜力更好。首次研究了模型自洽系统对GNPs辛醇/水分配系数的建模。

图形摘要

更新日期:2021-06-05
down
wechat
bug