A probabilistic approach to classifying metabolic stability

Anton Schwaighofer, Timon Schroeter, Sebastian Mika, Katja Hansen, Antonius Ter Laak, Philip Lienau, Andreas Reichel, Nikolaus Heinrich, Klaus Muller

Research output: Contribution to journalArticle

28 Citations (Scopus)

Abstract

Metabolie stability is an important property of drug molecules that should - optimally - be taken into account early on in the drug design process. Along with numerous medium- or high-throughput assays being implemented in early drug discovery, a prediction tool for this property could be of high value. However, metabolic stability is inherently difficult to predict, and no commercial tools are available for this purpose. In this work, we present a machine learning approach to predicting metabolic stability that is tailored to compounds from the drug development process at Bayer Schering Pharma. For four different in vitro assays, we develop Bayesian classification models to predict the probability of a compound being metabolically stable. The chosen approach implicitly takes the "domain of applicability" into account. The developed models were validated on recent project data at Bayer Schering Pharma, showing that the predictions are highly accurate and the domain of applicability is estimated correctly. Furthermore, we evaluate the modeling method on a set of publicly available data.

Original languageEnglish
Pages (from-to)785-796
Number of pages12
JournalJournal of Chemical Information and Modeling
Volume48
Issue number4
DOIs
Publication statusPublished - 2008 Apr 1
Externally publishedYes

Fingerprint

drug
Assays
Pharmaceutical Preparations
Learning systems
Throughput
Molecules
learning
Drug Discovery

ASJC Scopus subject areas

  • Chemistry(all)
  • Computational Theory and Mathematics
  • Computer Science Applications
  • Information Systems

Cite this

Schwaighofer, A., Schroeter, T., Mika, S., Hansen, K., Ter Laak, A., Lienau, P., ... Muller, K. (2008). A probabilistic approach to classifying metabolic stability. Journal of Chemical Information and Modeling, 48(4), 785-796. https://doi.org/10.1021/ci700142c

A probabilistic approach to classifying metabolic stability. / Schwaighofer, Anton; Schroeter, Timon; Mika, Sebastian; Hansen, Katja; Ter Laak, Antonius; Lienau, Philip; Reichel, Andreas; Heinrich, Nikolaus; Muller, Klaus.

In: Journal of Chemical Information and Modeling, Vol. 48, No. 4, 01.04.2008, p. 785-796.

Research output: Contribution to journalArticle

Schwaighofer, A, Schroeter, T, Mika, S, Hansen, K, Ter Laak, A, Lienau, P, Reichel, A, Heinrich, N & Muller, K 2008, 'A probabilistic approach to classifying metabolic stability', Journal of Chemical Information and Modeling, vol. 48, no. 4, pp. 785-796. https://doi.org/10.1021/ci700142c
Schwaighofer A, Schroeter T, Mika S, Hansen K, Ter Laak A, Lienau P et al. A probabilistic approach to classifying metabolic stability. Journal of Chemical Information and Modeling. 2008 Apr 1;48(4):785-796. https://doi.org/10.1021/ci700142c
Schwaighofer, Anton ; Schroeter, Timon ; Mika, Sebastian ; Hansen, Katja ; Ter Laak, Antonius ; Lienau, Philip ; Reichel, Andreas ; Heinrich, Nikolaus ; Muller, Klaus. / A probabilistic approach to classifying metabolic stability. In: Journal of Chemical Information and Modeling. 2008 ; Vol. 48, No. 4. pp. 785-796.
@article{65e3a175df4a4ab98763e59fa5b3404d,
title = "A probabilistic approach to classifying metabolic stability",
abstract = "Metabolie stability is an important property of drug molecules that should - optimally - be taken into account early on in the drug design process. Along with numerous medium- or high-throughput assays being implemented in early drug discovery, a prediction tool for this property could be of high value. However, metabolic stability is inherently difficult to predict, and no commercial tools are available for this purpose. In this work, we present a machine learning approach to predicting metabolic stability that is tailored to compounds from the drug development process at Bayer Schering Pharma. For four different in vitro assays, we develop Bayesian classification models to predict the probability of a compound being metabolically stable. The chosen approach implicitly takes the {"}domain of applicability{"} into account. The developed models were validated on recent project data at Bayer Schering Pharma, showing that the predictions are highly accurate and the domain of applicability is estimated correctly. Furthermore, we evaluate the modeling method on a set of publicly available data.",
author = "Anton Schwaighofer and Timon Schroeter and Sebastian Mika and Katja Hansen and {Ter Laak}, Antonius and Philip Lienau and Andreas Reichel and Nikolaus Heinrich and Klaus Muller",
year = "2008",
month = "4",
day = "1",
doi = "10.1021/ci700142c",
language = "English",
volume = "48",
pages = "785--796",
journal = "Journal of Chemical Information and Computer Sciences",
issn = "0095-2338",
publisher = "American Chemical Society",
number = "4",

}

TY - JOUR

T1 - A probabilistic approach to classifying metabolic stability

AU - Schwaighofer, Anton

AU - Schroeter, Timon

AU - Mika, Sebastian

AU - Hansen, Katja

AU - Ter Laak, Antonius

AU - Lienau, Philip

AU - Reichel, Andreas

AU - Heinrich, Nikolaus

AU - Muller, Klaus

PY - 2008/4/1

Y1 - 2008/4/1

N2 - Metabolie stability is an important property of drug molecules that should - optimally - be taken into account early on in the drug design process. Along with numerous medium- or high-throughput assays being implemented in early drug discovery, a prediction tool for this property could be of high value. However, metabolic stability is inherently difficult to predict, and no commercial tools are available for this purpose. In this work, we present a machine learning approach to predicting metabolic stability that is tailored to compounds from the drug development process at Bayer Schering Pharma. For four different in vitro assays, we develop Bayesian classification models to predict the probability of a compound being metabolically stable. The chosen approach implicitly takes the "domain of applicability" into account. The developed models were validated on recent project data at Bayer Schering Pharma, showing that the predictions are highly accurate and the domain of applicability is estimated correctly. Furthermore, we evaluate the modeling method on a set of publicly available data.

AB - Metabolie stability is an important property of drug molecules that should - optimally - be taken into account early on in the drug design process. Along with numerous medium- or high-throughput assays being implemented in early drug discovery, a prediction tool for this property could be of high value. However, metabolic stability is inherently difficult to predict, and no commercial tools are available for this purpose. In this work, we present a machine learning approach to predicting metabolic stability that is tailored to compounds from the drug development process at Bayer Schering Pharma. For four different in vitro assays, we develop Bayesian classification models to predict the probability of a compound being metabolically stable. The chosen approach implicitly takes the "domain of applicability" into account. The developed models were validated on recent project data at Bayer Schering Pharma, showing that the predictions are highly accurate and the domain of applicability is estimated correctly. Furthermore, we evaluate the modeling method on a set of publicly available data.

UR - http://www.scopus.com/inward/record.url?scp=44449142895&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=44449142895&partnerID=8YFLogxK

U2 - 10.1021/ci700142c

DO - 10.1021/ci700142c

M3 - Article

C2 - 18327900

AN - SCOPUS:44449142895

VL - 48

SP - 785

EP - 796

JO - Journal of Chemical Information and Computer Sciences

JF - Journal of Chemical Information and Computer Sciences

SN - 0095-2338

IS - 4

ER -