Nonlinear mixed effects dose repsonse modeling in high throughput drug screens: Application to melanoma cell line analysis

Kuan Fu Ding, Emanuel F. Petricoin, Darren Finlay, Hongwei Yin, William P.D. Hendricks, Chris Sereduk, Jeffrey Kiefer, Aleksandar D Sekulic, Patricia M. LoRusso, Kristiina Vuori, Jeffrey M. Trent, Nicholas J. Schork

Research output: Contribution to journalArticle

Abstract

Cancer cell lines are often used in high throughput drug screens (HTS) to explore the relationship between cell line characteristics and responsiveness to different therapies. Many current analysis methods infer relationships by focusing on one aspect of cell line drug-specific dose-response curves (DRCs), the concentration causing 50% inhibition of a phenotypic endpoint (IC50). Such methods may overlook DRC features and do not simultaneously leverage information about drug response patterns across cell lines, potentially increasing false positive and negative rates in drug response associations. We consider the application of two methods, each rooted in nonlinear mixed effects (NLME) models, that test the relationship relationships between estimated cell line DRCs and factors that might mitigate response. Both methods leverage estimation and testing techniques that consider the simultaneous analysis of different cell lines to draw inferences about any one cell line. One of the methods is designed to provide an omnibus test of the differences between cell line DRCs that is not focused on any one aspect of the DRC (such as the IC50 value). We simulated different settings and compared the different methods on the simulated data. We also compared the proposed methods against traditional IC50-based methods using 40 melanoma cell lines whose transcriptomes, proteomes, and, importantly, BRAF and related mutation profiles were available. Ultimately, we find that the NLMEbased methods are more robust, powerful and, for the omnibus test, more flexible, than traditional methods. Their application to the melanoma cell lines reveals insights into factors that may be clinically useful.

Original languageEnglish (US)
Pages (from-to)5044-5057
Number of pages14
JournalOncotarget
Volume9
Issue number4
DOIs
StatePublished - Jan 1 2018
Externally publishedYes

Fingerprint

Melanoma
Cell Line
Pharmaceutical Preparations
Inhibitory Concentration 50
Proteome
Transcriptome
Mutation

Keywords

  • Bioinformatics
  • Cancer
  • Drug response
  • High throughput drug screen
  • Nonlinear mixed effect models

ASJC Scopus subject areas

  • Oncology

Cite this

Ding, K. F., Petricoin, E. F., Finlay, D., Yin, H., Hendricks, W. P. D., Sereduk, C., ... Schork, N. J. (2018). Nonlinear mixed effects dose repsonse modeling in high throughput drug screens: Application to melanoma cell line analysis. Oncotarget, 9(4), 5044-5057. https://doi.org/10.18632/oncotarget.23495

Nonlinear mixed effects dose repsonse modeling in high throughput drug screens : Application to melanoma cell line analysis. / Ding, Kuan Fu; Petricoin, Emanuel F.; Finlay, Darren; Yin, Hongwei; Hendricks, William P.D.; Sereduk, Chris; Kiefer, Jeffrey; Sekulic, Aleksandar D; LoRusso, Patricia M.; Vuori, Kristiina; Trent, Jeffrey M.; Schork, Nicholas J.

In: Oncotarget, Vol. 9, No. 4, 01.01.2018, p. 5044-5057.

Research output: Contribution to journalArticle

Ding, KF, Petricoin, EF, Finlay, D, Yin, H, Hendricks, WPD, Sereduk, C, Kiefer, J, Sekulic, AD, LoRusso, PM, Vuori, K, Trent, JM & Schork, NJ 2018, 'Nonlinear mixed effects dose repsonse modeling in high throughput drug screens: Application to melanoma cell line analysis', Oncotarget, vol. 9, no. 4, pp. 5044-5057. https://doi.org/10.18632/oncotarget.23495
Ding, Kuan Fu ; Petricoin, Emanuel F. ; Finlay, Darren ; Yin, Hongwei ; Hendricks, William P.D. ; Sereduk, Chris ; Kiefer, Jeffrey ; Sekulic, Aleksandar D ; LoRusso, Patricia M. ; Vuori, Kristiina ; Trent, Jeffrey M. ; Schork, Nicholas J. / Nonlinear mixed effects dose repsonse modeling in high throughput drug screens : Application to melanoma cell line analysis. In: Oncotarget. 2018 ; Vol. 9, No. 4. pp. 5044-5057.
@article{98dd65b1f1be46949b7511c33e5bb76c,
title = "Nonlinear mixed effects dose repsonse modeling in high throughput drug screens: Application to melanoma cell line analysis",
abstract = "Cancer cell lines are often used in high throughput drug screens (HTS) to explore the relationship between cell line characteristics and responsiveness to different therapies. Many current analysis methods infer relationships by focusing on one aspect of cell line drug-specific dose-response curves (DRCs), the concentration causing 50{\%} inhibition of a phenotypic endpoint (IC50). Such methods may overlook DRC features and do not simultaneously leverage information about drug response patterns across cell lines, potentially increasing false positive and negative rates in drug response associations. We consider the application of two methods, each rooted in nonlinear mixed effects (NLME) models, that test the relationship relationships between estimated cell line DRCs and factors that might mitigate response. Both methods leverage estimation and testing techniques that consider the simultaneous analysis of different cell lines to draw inferences about any one cell line. One of the methods is designed to provide an omnibus test of the differences between cell line DRCs that is not focused on any one aspect of the DRC (such as the IC50 value). We simulated different settings and compared the different methods on the simulated data. We also compared the proposed methods against traditional IC50-based methods using 40 melanoma cell lines whose transcriptomes, proteomes, and, importantly, BRAF and related mutation profiles were available. Ultimately, we find that the NLMEbased methods are more robust, powerful and, for the omnibus test, more flexible, than traditional methods. Their application to the melanoma cell lines reveals insights into factors that may be clinically useful.",
keywords = "Bioinformatics, Cancer, Drug response, High throughput drug screen, Nonlinear mixed effect models",
author = "Ding, {Kuan Fu} and Petricoin, {Emanuel F.} and Darren Finlay and Hongwei Yin and Hendricks, {William P.D.} and Chris Sereduk and Jeffrey Kiefer and Sekulic, {Aleksandar D} and LoRusso, {Patricia M.} and Kristiina Vuori and Trent, {Jeffrey M.} and Schork, {Nicholas J.}",
year = "2018",
month = "1",
day = "1",
doi = "10.18632/oncotarget.23495",
language = "English (US)",
volume = "9",
pages = "5044--5057",
journal = "Oncotarget",
issn = "1949-2553",
publisher = "Impact Journals",
number = "4",

}

TY - JOUR

T1 - Nonlinear mixed effects dose repsonse modeling in high throughput drug screens

T2 - Application to melanoma cell line analysis

AU - Ding, Kuan Fu

AU - Petricoin, Emanuel F.

AU - Finlay, Darren

AU - Yin, Hongwei

AU - Hendricks, William P.D.

AU - Sereduk, Chris

AU - Kiefer, Jeffrey

AU - Sekulic, Aleksandar D

AU - LoRusso, Patricia M.

AU - Vuori, Kristiina

AU - Trent, Jeffrey M.

AU - Schork, Nicholas J.

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Cancer cell lines are often used in high throughput drug screens (HTS) to explore the relationship between cell line characteristics and responsiveness to different therapies. Many current analysis methods infer relationships by focusing on one aspect of cell line drug-specific dose-response curves (DRCs), the concentration causing 50% inhibition of a phenotypic endpoint (IC50). Such methods may overlook DRC features and do not simultaneously leverage information about drug response patterns across cell lines, potentially increasing false positive and negative rates in drug response associations. We consider the application of two methods, each rooted in nonlinear mixed effects (NLME) models, that test the relationship relationships between estimated cell line DRCs and factors that might mitigate response. Both methods leverage estimation and testing techniques that consider the simultaneous analysis of different cell lines to draw inferences about any one cell line. One of the methods is designed to provide an omnibus test of the differences between cell line DRCs that is not focused on any one aspect of the DRC (such as the IC50 value). We simulated different settings and compared the different methods on the simulated data. We also compared the proposed methods against traditional IC50-based methods using 40 melanoma cell lines whose transcriptomes, proteomes, and, importantly, BRAF and related mutation profiles were available. Ultimately, we find that the NLMEbased methods are more robust, powerful and, for the omnibus test, more flexible, than traditional methods. Their application to the melanoma cell lines reveals insights into factors that may be clinically useful.

AB - Cancer cell lines are often used in high throughput drug screens (HTS) to explore the relationship between cell line characteristics and responsiveness to different therapies. Many current analysis methods infer relationships by focusing on one aspect of cell line drug-specific dose-response curves (DRCs), the concentration causing 50% inhibition of a phenotypic endpoint (IC50). Such methods may overlook DRC features and do not simultaneously leverage information about drug response patterns across cell lines, potentially increasing false positive and negative rates in drug response associations. We consider the application of two methods, each rooted in nonlinear mixed effects (NLME) models, that test the relationship relationships between estimated cell line DRCs and factors that might mitigate response. Both methods leverage estimation and testing techniques that consider the simultaneous analysis of different cell lines to draw inferences about any one cell line. One of the methods is designed to provide an omnibus test of the differences between cell line DRCs that is not focused on any one aspect of the DRC (such as the IC50 value). We simulated different settings and compared the different methods on the simulated data. We also compared the proposed methods against traditional IC50-based methods using 40 melanoma cell lines whose transcriptomes, proteomes, and, importantly, BRAF and related mutation profiles were available. Ultimately, we find that the NLMEbased methods are more robust, powerful and, for the omnibus test, more flexible, than traditional methods. Their application to the melanoma cell lines reveals insights into factors that may be clinically useful.

KW - Bioinformatics

KW - Cancer

KW - Drug response

KW - High throughput drug screen

KW - Nonlinear mixed effect models

UR - http://www.scopus.com/inward/record.url?scp=85040444794&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85040444794&partnerID=8YFLogxK

U2 - 10.18632/oncotarget.23495

DO - 10.18632/oncotarget.23495

M3 - Article

AN - SCOPUS:85040444794

VL - 9

SP - 5044

EP - 5057

JO - Oncotarget

JF - Oncotarget

SN - 1949-2553

IS - 4

ER -