2022
DOI: 10.1021/acs.analchem.2c01667
|View full text |Cite
|
Sign up to set email alerts
|

Direct Prediction of Physicochemical Properties and Toxicities of Chemicals from Analytical Descriptors by GC–MS

Abstract: With advances in machine learning (ML) techniques, the quantitative structure–activity relationship (QSAR) approach is becoming popular for evaluating chemicals. However, the QSAR approach requires that the chemical structure of the target compound is known and that it should be convertible to molecular descriptors. These requirements lead to limitations in predicting the properties and toxicities of chemicals distributed in the environment as in the PubChem database; the structural information on only 14% of … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
2

Citation Types

0
8
0

Year Published

2023
2023
2025
2025

Publication Types

Select...
6
1

Relationship

1
6

Authors

Journals

citations
Cited by 9 publications
(8 citation statements)
references
References 50 publications
0
8
0
Order By: Relevance
“…Datasets of analytical and molecular descriptors were also used to build a QSAR prediction model based on XGBoost [28]. Prediction models using each molecular descriptor as input were prepared in this study and compared to Detective-QSAR, that uses the analytical descriptor as input [9]. The analytical descriptor was not standardized for construction of Detective-QSAR, unlike for t-SNE, because standardization does not affect the calculation results obtained by XGBoost, unlike those acquired by t-SNE.…”
Section: ∂L ∂Ymentioning
confidence: 99%
See 2 more Smart Citations
“…Datasets of analytical and molecular descriptors were also used to build a QSAR prediction model based on XGBoost [28]. Prediction models using each molecular descriptor as input were prepared in this study and compared to Detective-QSAR, that uses the analytical descriptor as input [9]. The analytical descriptor was not standardized for construction of Detective-QSAR, unlike for t-SNE, because standardization does not affect the calculation results obtained by XGBoost, unlike those acquired by t-SNE.…”
Section: ∂L ∂Ymentioning
confidence: 99%
“…The training dataset was consistent among all models with the same subject, except for a few cases of data removal because of incomplete data. Model performance was evaluated using a combination of the validation and test datasets because the validation dataset did not have a remarkable in uence on model performance in the previous study [9].…”
Section: ∂L ∂Ymentioning
confidence: 99%
See 1 more Smart Citation
“…Such tools can rely on biochemical laboratory approaches [8][9][10] or machine-learning algorithms trained on existing toxicity databases. [11][12][13] Because establishing toxicity is exceptionally difficult, a complementary approach might seek to identify chemical features with the potential to cause high exposure in humans and/or wildlife.…”
Section: Introductionmentioning
confidence: 99%
“…Chromatographic retention times and mass spectrometric fragmentation spectra (MS 2 ) that can be obtained from NTA contain information relevant for the chemical properties controlling exposure, 19 with a recent study applying machine learning to predict, for example, vapour pressure and the equilibrium octanol-water partitioning ratio K OW of chemical features. 11 Most NTA yields information on the accurate molecular mass of a chemical feature, from which the molecular formula is relatively easy to derive compared to assigning a molecular structure, especially when these compounds are outside the current chemical libraries. 20 Even if the molecular structures of some features could be assigned by matching to chemical libraries, relatively high false discoveries might limit structure-based chemical prioritization.…”
Section: Introductionmentioning
confidence: 99%