University of Hertfordshire

From the same journal

By the same authors

Standard

Development of a Gaussian Process – Feature Selection Model to Characterise (poly)dimethylsiloxane (Silastic®) Membrane Permeation. / Sun, Yi; Hewitt, Mark; Wilkinson, Simon C; Davey, Neil; Adams, Roderick; Gullick, Darren ; Moss, Gary.

In: Journal of Pharmacy and Pharmacology, Vol. 72, No. 7, 07.2020, p. 873-888.

Research output: Contribution to journalArticlepeer-review

Harvard

APA

Vancouver

Author

Bibtex

@article{79fcaaa798cc4e0fb523e209998866c6,
title = "Development of a Gaussian Process – Feature Selection Model to Characterise (poly)dimethylsiloxane (Silastic{\textregistered}) Membrane Permeation",
abstract = "ObjectivesThe current study aims to determine the effect of physicochemical descriptor selection on models of polydimethylsiloxane permeation.Methods2,942 descriptors were calculated for a dataset of 77 chemicals. Data was processed to remove redundancy, single values, imbalanced and highly correlated data, yielding 1,363 relevant descriptors. For four independent test sets feature selection methods were applied and modelled via a variety of Machine Learning methods. Key findingsTwo sets of molecular descriptors which can provide improved predictions, compared to existing models, have been identified. Best permeation predictions were found with Gaussian Process methods. The molecular descriptors describe lipophilicity, partial charge and hydrogen bonding as key determinants of PDMS permeation. ConclusionsThis study highlights important considerations in the development of relevant models and in the construction and use of the datasets used in such studies, particularly that highly correlated descriptors should be removed from datasets. Predictive models are improved by the methodology adopted in this study, notably the systematic evaluation of descriptors, rather than simply using any and all available descriptors, often based empirically on in vitro experiments. Such findings also have clear relevance to a number of other fields",
author = "Yi Sun and Mark Hewitt and Wilkinson, {Simon C} and Neil Davey and Roderick Adams and Darren Gullick and Gary Moss",
note = "{\textcopyright} 2020 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology. This is an open access article under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits use, distribution and reproduction in any medium, provided the original work is properly cited.",
year = "2020",
month = jul,
doi = "10.1111/jphp.13263",
language = "English",
volume = "72",
pages = "873--888",
journal = "Journal of Pharmacy and Pharmacology",
issn = "0022-3573",
publisher = "Pharmaceutical Press",
number = "7",

}

RIS

TY - JOUR

T1 - Development of a Gaussian Process – Feature Selection Model to Characterise (poly)dimethylsiloxane (Silastic®) Membrane Permeation

AU - Sun, Yi

AU - Hewitt, Mark

AU - Wilkinson, Simon C

AU - Davey, Neil

AU - Adams, Roderick

AU - Gullick, Darren

AU - Moss, Gary

N1 - © 2020 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology. This is an open access article under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

PY - 2020/7

Y1 - 2020/7

N2 - ObjectivesThe current study aims to determine the effect of physicochemical descriptor selection on models of polydimethylsiloxane permeation.Methods2,942 descriptors were calculated for a dataset of 77 chemicals. Data was processed to remove redundancy, single values, imbalanced and highly correlated data, yielding 1,363 relevant descriptors. For four independent test sets feature selection methods were applied and modelled via a variety of Machine Learning methods. Key findingsTwo sets of molecular descriptors which can provide improved predictions, compared to existing models, have been identified. Best permeation predictions were found with Gaussian Process methods. The molecular descriptors describe lipophilicity, partial charge and hydrogen bonding as key determinants of PDMS permeation. ConclusionsThis study highlights important considerations in the development of relevant models and in the construction and use of the datasets used in such studies, particularly that highly correlated descriptors should be removed from datasets. Predictive models are improved by the methodology adopted in this study, notably the systematic evaluation of descriptors, rather than simply using any and all available descriptors, often based empirically on in vitro experiments. Such findings also have clear relevance to a number of other fields

AB - ObjectivesThe current study aims to determine the effect of physicochemical descriptor selection on models of polydimethylsiloxane permeation.Methods2,942 descriptors were calculated for a dataset of 77 chemicals. Data was processed to remove redundancy, single values, imbalanced and highly correlated data, yielding 1,363 relevant descriptors. For four independent test sets feature selection methods were applied and modelled via a variety of Machine Learning methods. Key findingsTwo sets of molecular descriptors which can provide improved predictions, compared to existing models, have been identified. Best permeation predictions were found with Gaussian Process methods. The molecular descriptors describe lipophilicity, partial charge and hydrogen bonding as key determinants of PDMS permeation. ConclusionsThis study highlights important considerations in the development of relevant models and in the construction and use of the datasets used in such studies, particularly that highly correlated descriptors should be removed from datasets. Predictive models are improved by the methodology adopted in this study, notably the systematic evaluation of descriptors, rather than simply using any and all available descriptors, often based empirically on in vitro experiments. Such findings also have clear relevance to a number of other fields

U2 - 10.1111/jphp.13263

DO - 10.1111/jphp.13263

M3 - Article

VL - 72

SP - 873

EP - 888

JO - Journal of Pharmacy and Pharmacology

JF - Journal of Pharmacy and Pharmacology

SN - 0022-3573

IS - 7

ER -