Machine learning approaches for predicting compounds that interact with therapeutic and ADMET related proteins

Hu Li, C. W. Yap, C. Y. Ung, Y. Xue, Z. R. Li, L. Y. Han, H. H. Lin, Y. Z. Chen

Research output: Contribution to journalArticle

46 Citations (Scopus)

Abstract

Computational methods for predicting compounds of specific pharmacodynamic and ADMET (absorption, distribution, metabolism, excretion and toxicity) property are useful for facilitating drug discovery and evaluation. Recently, machine learning methods such as neural networks and support vector machines have been explored for predicting inhibitors, antagonists, blockers, agonists, activators and substrates of proteins related to specific therapeutic and ADMET property. These methods are particularly useful for compounds of diverse structures to complement QSAR methods, and for cases of unavailable receptor 3D structure to complement structure-based methods. A number of studies have demonstrated the potential of these methods for predicting such compounds as substrates of P-glycoprotein and cytochrome P450 CYP isoenzymes, inhibitors of protein kinases and CYP isoenzymes, and agonists of serotonin receptor and estrogen receptor. This article is intended to review the strategies, current progresses and underlying difficulties in using machine learning methods for predicting these protein binders and as potential virtual screening tools. Algorithms for proper representation of the structural and physicochemical properties of compounds are also evaluated.

Original languageEnglish (US)
Pages (from-to)2838-2860
Number of pages23
JournalJournal of Pharmaceutical Sciences
Volume96
Issue number11
DOIs
StatePublished - Nov 2007
Externally publishedYes

Fingerprint

Metabolism
Isoenzymes
Toxicity
Learning systems
Pharmacodynamics
Serotonin Receptor Agonists
P-Glycoprotein
Substrates
Protein Kinase Inhibitors
Computational methods
Estrogen Receptors
Cytochrome P-450 Enzyme System
Binders
Support vector machines
Screening
Proteins
Neural networks
Therapeutics
Drug Evaluation
Quantitative Structure-Activity Relationship

Keywords

  • Computer aided drug design
  • High throughput technologies
  • In silico modeling
  • Neural networks
  • Pharmacokinetic/pharmacodynamic models

ASJC Scopus subject areas

  • Drug Discovery
  • Organic Chemistry
  • Chemistry(all)
  • Molecular Medicine
  • Pharmacology
  • Pharmaceutical Science

Cite this

Machine learning approaches for predicting compounds that interact with therapeutic and ADMET related proteins. / Li, Hu; Yap, C. W.; Ung, C. Y.; Xue, Y.; Li, Z. R.; Han, L. Y.; Lin, H. H.; Chen, Y. Z.

In: Journal of Pharmaceutical Sciences, Vol. 96, No. 11, 11.2007, p. 2838-2860.

Research output: Contribution to journalArticle

Li, Hu ; Yap, C. W. ; Ung, C. Y. ; Xue, Y. ; Li, Z. R. ; Han, L. Y. ; Lin, H. H. ; Chen, Y. Z. / Machine learning approaches for predicting compounds that interact with therapeutic and ADMET related proteins. In: Journal of Pharmaceutical Sciences. 2007 ; Vol. 96, No. 11. pp. 2838-2860.
@article{604cd1e3b9b14e1090d768334bd217c3,
title = "Machine learning approaches for predicting compounds that interact with therapeutic and ADMET related proteins",
abstract = "Computational methods for predicting compounds of specific pharmacodynamic and ADMET (absorption, distribution, metabolism, excretion and toxicity) property are useful for facilitating drug discovery and evaluation. Recently, machine learning methods such as neural networks and support vector machines have been explored for predicting inhibitors, antagonists, blockers, agonists, activators and substrates of proteins related to specific therapeutic and ADMET property. These methods are particularly useful for compounds of diverse structures to complement QSAR methods, and for cases of unavailable receptor 3D structure to complement structure-based methods. A number of studies have demonstrated the potential of these methods for predicting such compounds as substrates of P-glycoprotein and cytochrome P450 CYP isoenzymes, inhibitors of protein kinases and CYP isoenzymes, and agonists of serotonin receptor and estrogen receptor. This article is intended to review the strategies, current progresses and underlying difficulties in using machine learning methods for predicting these protein binders and as potential virtual screening tools. Algorithms for proper representation of the structural and physicochemical properties of compounds are also evaluated.",
keywords = "Computer aided drug design, High throughput technologies, In silico modeling, Neural networks, Pharmacokinetic/pharmacodynamic models",
author = "Hu Li and Yap, {C. W.} and Ung, {C. Y.} and Y. Xue and Li, {Z. R.} and Han, {L. Y.} and Lin, {H. H.} and Chen, {Y. Z.}",
year = "2007",
month = "11",
doi = "10.1002/jps.20985",
language = "English (US)",
volume = "96",
pages = "2838--2860",
journal = "Journal of Pharmaceutical Sciences",
issn = "0022-3549",
publisher = "John Wiley and Sons Inc.",
number = "11",

}

TY - JOUR

T1 - Machine learning approaches for predicting compounds that interact with therapeutic and ADMET related proteins

AU - Li, Hu

AU - Yap, C. W.

AU - Ung, C. Y.

AU - Xue, Y.

AU - Li, Z. R.

AU - Han, L. Y.

AU - Lin, H. H.

AU - Chen, Y. Z.

PY - 2007/11

Y1 - 2007/11

N2 - Computational methods for predicting compounds of specific pharmacodynamic and ADMET (absorption, distribution, metabolism, excretion and toxicity) property are useful for facilitating drug discovery and evaluation. Recently, machine learning methods such as neural networks and support vector machines have been explored for predicting inhibitors, antagonists, blockers, agonists, activators and substrates of proteins related to specific therapeutic and ADMET property. These methods are particularly useful for compounds of diverse structures to complement QSAR methods, and for cases of unavailable receptor 3D structure to complement structure-based methods. A number of studies have demonstrated the potential of these methods for predicting such compounds as substrates of P-glycoprotein and cytochrome P450 CYP isoenzymes, inhibitors of protein kinases and CYP isoenzymes, and agonists of serotonin receptor and estrogen receptor. This article is intended to review the strategies, current progresses and underlying difficulties in using machine learning methods for predicting these protein binders and as potential virtual screening tools. Algorithms for proper representation of the structural and physicochemical properties of compounds are also evaluated.

AB - Computational methods for predicting compounds of specific pharmacodynamic and ADMET (absorption, distribution, metabolism, excretion and toxicity) property are useful for facilitating drug discovery and evaluation. Recently, machine learning methods such as neural networks and support vector machines have been explored for predicting inhibitors, antagonists, blockers, agonists, activators and substrates of proteins related to specific therapeutic and ADMET property. These methods are particularly useful for compounds of diverse structures to complement QSAR methods, and for cases of unavailable receptor 3D structure to complement structure-based methods. A number of studies have demonstrated the potential of these methods for predicting such compounds as substrates of P-glycoprotein and cytochrome P450 CYP isoenzymes, inhibitors of protein kinases and CYP isoenzymes, and agonists of serotonin receptor and estrogen receptor. This article is intended to review the strategies, current progresses and underlying difficulties in using machine learning methods for predicting these protein binders and as potential virtual screening tools. Algorithms for proper representation of the structural and physicochemical properties of compounds are also evaluated.

KW - Computer aided drug design

KW - High throughput technologies

KW - In silico modeling

KW - Neural networks

KW - Pharmacokinetic/pharmacodynamic models

UR - http://www.scopus.com/inward/record.url?scp=36849009228&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=36849009228&partnerID=8YFLogxK

U2 - 10.1002/jps.20985

DO - 10.1002/jps.20985

M3 - Article

C2 - 17786989

AN - SCOPUS:36849009228

VL - 96

SP - 2838

EP - 2860

JO - Journal of Pharmaceutical Sciences

JF - Journal of Pharmaceutical Sciences

SN - 0022-3549

IS - 11

ER -