Utilizing high throughput screening data for predictive toxicology models: Protocols and application to MLSCN assays

Rajarshi Guha, Stephan C Schuerer

Research output: Contribution to journalArticle

42 Citations (Scopus)

Abstract

Computational toxicology is emerging as an encouraging alternative to experimental testing. The Molecular Libraries Screening Center Network (MLSCN) as part of the NIH Molecular Libraries Roadmap has recently started generating large and diverse screening datasets, which are publicly available in PubChem. In this report, we investigate various aspects of developing computational models to predict cell toxicity based on cell proliferation screening data generated in the MLSCN. By capturing feature-based information in those datasets, such predictive models would be useful in evaluating cell-based screening results in general (for example from reporter assays) and could be used as an aid to identify and eliminate potentially undesired compounds. Specifically we present the results of random forest ensemble models developed using different cell proliferation datasets and highlight protocols to take into account their extremely imbalanced nature. Depending on the nature of the datasets and the descriptors employed we were able to achieve percentage correct classification rates between 70% and 85% on the prediction set, though the accuracy rate dropped significantly when the models were applied to in vivo data. In this context we also compare the MLSCN cell proliferation results with animal acute toxicity data to investigate to what extent animal toxicity can be correlated and potentially predicted by proliferation results. Finally, we present a visualization technique that allows one to compare a new dataset to the training set of the models to decide whether the new dataset may be reliably predicted.

Original languageEnglish
Pages (from-to)367-384
Number of pages18
JournalJournal of Computer-Aided Molecular Design
Volume22
Issue number6-7
DOIs
StatePublished - Jun 1 2008
Externally publishedYes

Fingerprint

toxicology
Toxicology
Assays
Screening
screening
Throughput
Cell proliferation
toxicity
Toxicity
Cell Proliferation
animals
Animals
cells
Datasets
emerging
education
Visualization
Testing
predictions

Keywords

  • Animal toxicity
  • Cell proliferation
  • Domain applicability
  • HT Sassay
  • Jurkat cell line
  • QSAR

ASJC Scopus subject areas

  • Molecular Medicine

Cite this

@article{060f6f850a174232b337195a9133a667,
title = "Utilizing high throughput screening data for predictive toxicology models: Protocols and application to MLSCN assays",
abstract = "Computational toxicology is emerging as an encouraging alternative to experimental testing. The Molecular Libraries Screening Center Network (MLSCN) as part of the NIH Molecular Libraries Roadmap has recently started generating large and diverse screening datasets, which are publicly available in PubChem. In this report, we investigate various aspects of developing computational models to predict cell toxicity based on cell proliferation screening data generated in the MLSCN. By capturing feature-based information in those datasets, such predictive models would be useful in evaluating cell-based screening results in general (for example from reporter assays) and could be used as an aid to identify and eliminate potentially undesired compounds. Specifically we present the results of random forest ensemble models developed using different cell proliferation datasets and highlight protocols to take into account their extremely imbalanced nature. Depending on the nature of the datasets and the descriptors employed we were able to achieve percentage correct classification rates between 70{\%} and 85{\%} on the prediction set, though the accuracy rate dropped significantly when the models were applied to in vivo data. In this context we also compare the MLSCN cell proliferation results with animal acute toxicity data to investigate to what extent animal toxicity can be correlated and potentially predicted by proliferation results. Finally, we present a visualization technique that allows one to compare a new dataset to the training set of the models to decide whether the new dataset may be reliably predicted.",
keywords = "Animal toxicity, Cell proliferation, Domain applicability, HT Sassay, Jurkat cell line, QSAR",
author = "Rajarshi Guha and Schuerer, {Stephan C}",
year = "2008",
month = "6",
day = "1",
doi = "10.1007/s10822-008-9192-9",
language = "English",
volume = "22",
pages = "367--384",
journal = "Journal of Computer-Aided Molecular Design",
issn = "0920-654X",
publisher = "Springer Netherlands",
number = "6-7",

}

TY - JOUR

T1 - Utilizing high throughput screening data for predictive toxicology models

T2 - Protocols and application to MLSCN assays

AU - Guha, Rajarshi

AU - Schuerer, Stephan C

PY - 2008/6/1

Y1 - 2008/6/1

N2 - Computational toxicology is emerging as an encouraging alternative to experimental testing. The Molecular Libraries Screening Center Network (MLSCN) as part of the NIH Molecular Libraries Roadmap has recently started generating large and diverse screening datasets, which are publicly available in PubChem. In this report, we investigate various aspects of developing computational models to predict cell toxicity based on cell proliferation screening data generated in the MLSCN. By capturing feature-based information in those datasets, such predictive models would be useful in evaluating cell-based screening results in general (for example from reporter assays) and could be used as an aid to identify and eliminate potentially undesired compounds. Specifically we present the results of random forest ensemble models developed using different cell proliferation datasets and highlight protocols to take into account their extremely imbalanced nature. Depending on the nature of the datasets and the descriptors employed we were able to achieve percentage correct classification rates between 70% and 85% on the prediction set, though the accuracy rate dropped significantly when the models were applied to in vivo data. In this context we also compare the MLSCN cell proliferation results with animal acute toxicity data to investigate to what extent animal toxicity can be correlated and potentially predicted by proliferation results. Finally, we present a visualization technique that allows one to compare a new dataset to the training set of the models to decide whether the new dataset may be reliably predicted.

AB - Computational toxicology is emerging as an encouraging alternative to experimental testing. The Molecular Libraries Screening Center Network (MLSCN) as part of the NIH Molecular Libraries Roadmap has recently started generating large and diverse screening datasets, which are publicly available in PubChem. In this report, we investigate various aspects of developing computational models to predict cell toxicity based on cell proliferation screening data generated in the MLSCN. By capturing feature-based information in those datasets, such predictive models would be useful in evaluating cell-based screening results in general (for example from reporter assays) and could be used as an aid to identify and eliminate potentially undesired compounds. Specifically we present the results of random forest ensemble models developed using different cell proliferation datasets and highlight protocols to take into account their extremely imbalanced nature. Depending on the nature of the datasets and the descriptors employed we were able to achieve percentage correct classification rates between 70% and 85% on the prediction set, though the accuracy rate dropped significantly when the models were applied to in vivo data. In this context we also compare the MLSCN cell proliferation results with animal acute toxicity data to investigate to what extent animal toxicity can be correlated and potentially predicted by proliferation results. Finally, we present a visualization technique that allows one to compare a new dataset to the training set of the models to decide whether the new dataset may be reliably predicted.

KW - Animal toxicity

KW - Cell proliferation

KW - Domain applicability

KW - HT Sassay

KW - Jurkat cell line

KW - QSAR

UR - http://www.scopus.com/inward/record.url?scp=44249108095&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=44249108095&partnerID=8YFLogxK

U2 - 10.1007/s10822-008-9192-9

DO - 10.1007/s10822-008-9192-9

M3 - Article

C2 - 18283419

AN - SCOPUS:44249108095

VL - 22

SP - 367

EP - 384

JO - Journal of Computer-Aided Molecular Design

JF - Journal of Computer-Aided Molecular Design

SN - 0920-654X

IS - 6-7

ER -