A General Framework for Formal Tests of Interaction after Exhaustive Search Methods with Applications to MDR and MDR-PDT

Todd L. Edwards, Stephen D. Turner, Eric S. Torstenson, Scott M. Dudek, Eden R. Martin, Marylyn Deriggi Ritchie

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

The initial presentation of multifactor dimensionality reduction (MDR) featured cross-validation to mitigate over-fitting, computationally efficient searches of the epistatic model space, and variable construction with constructive induction to alleviate the curse of dimensionality. However, the method was unable to differentiate association signals arising from true interactions from those due to independent main effects at individual loci. This issue leads to problems in inference and interpretability for the results from MDR and the family-based compliment the MDR-pedigree disequilibrium test (PDT). A suggestion from previous work was to fit regression models post hoc to specifically evaluate the null hypothesis of no interaction for MDR or MDR-PDT models. We demonstrate with simulation that fitting a regression model on the same data as that analyzed by MDR or MDR-PDT is not a valid test of interaction. This is likely to be true for any other procedure that searches for models, and then performs an uncorrected test for interaction. We also show with simulation that when strong main effects are present and the null hypothesis of no interaction is true, that MDR and MDR-PDT reject at far greater than the nominal rate. We also provide a valid regression-based permutation test procedure that specifically tests the null hypothesis of no interaction, and does not reject the null when only main effects are present. The regression-based permutation test implemented here conducts a valid test of interaction after a search for multilocus models, and can be applied to any method that conducts a search to find a multilocus model representing an interaction.

Original languageEnglish (US)
Article numbere9363
JournalPloS one
Volume5
Issue number2
DOIs
StatePublished - Feb 23 2010

Fingerprint

Multifactor Dimensionality Reduction
Pedigree
pedigree
testing
methodology
Space Simulation
Association reactions

All Science Journal Classification (ASJC) codes

  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)

Cite this

Edwards, T. L., Turner, S. D., Torstenson, E. S., Dudek, S. M., Martin, E. R., & Ritchie, M. D. (2010). A General Framework for Formal Tests of Interaction after Exhaustive Search Methods with Applications to MDR and MDR-PDT. PloS one, 5(2), [e9363]. https://doi.org/10.1371/journal.pone.0009363
Edwards, Todd L. ; Turner, Stephen D. ; Torstenson, Eric S. ; Dudek, Scott M. ; Martin, Eden R. ; Ritchie, Marylyn Deriggi. / A General Framework for Formal Tests of Interaction after Exhaustive Search Methods with Applications to MDR and MDR-PDT. In: PloS one. 2010 ; Vol. 5, No. 2.
@article{13c2db8283e24d52ab2694e1c85f9700,
title = "A General Framework for Formal Tests of Interaction after Exhaustive Search Methods with Applications to MDR and MDR-PDT",
abstract = "The initial presentation of multifactor dimensionality reduction (MDR) featured cross-validation to mitigate over-fitting, computationally efficient searches of the epistatic model space, and variable construction with constructive induction to alleviate the curse of dimensionality. However, the method was unable to differentiate association signals arising from true interactions from those due to independent main effects at individual loci. This issue leads to problems in inference and interpretability for the results from MDR and the family-based compliment the MDR-pedigree disequilibrium test (PDT). A suggestion from previous work was to fit regression models post hoc to specifically evaluate the null hypothesis of no interaction for MDR or MDR-PDT models. We demonstrate with simulation that fitting a regression model on the same data as that analyzed by MDR or MDR-PDT is not a valid test of interaction. This is likely to be true for any other procedure that searches for models, and then performs an uncorrected test for interaction. We also show with simulation that when strong main effects are present and the null hypothesis of no interaction is true, that MDR and MDR-PDT reject at far greater than the nominal rate. We also provide a valid regression-based permutation test procedure that specifically tests the null hypothesis of no interaction, and does not reject the null when only main effects are present. The regression-based permutation test implemented here conducts a valid test of interaction after a search for multilocus models, and can be applied to any method that conducts a search to find a multilocus model representing an interaction.",
author = "Edwards, {Todd L.} and Turner, {Stephen D.} and Torstenson, {Eric S.} and Dudek, {Scott M.} and Martin, {Eden R.} and Ritchie, {Marylyn Deriggi}",
year = "2010",
month = "2",
day = "23",
doi = "10.1371/journal.pone.0009363",
language = "English (US)",
volume = "5",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "2",

}

A General Framework for Formal Tests of Interaction after Exhaustive Search Methods with Applications to MDR and MDR-PDT. / Edwards, Todd L.; Turner, Stephen D.; Torstenson, Eric S.; Dudek, Scott M.; Martin, Eden R.; Ritchie, Marylyn Deriggi.

In: PloS one, Vol. 5, No. 2, e9363, 23.02.2010.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A General Framework for Formal Tests of Interaction after Exhaustive Search Methods with Applications to MDR and MDR-PDT

AU - Edwards, Todd L.

AU - Turner, Stephen D.

AU - Torstenson, Eric S.

AU - Dudek, Scott M.

AU - Martin, Eden R.

AU - Ritchie, Marylyn Deriggi

PY - 2010/2/23

Y1 - 2010/2/23

N2 - The initial presentation of multifactor dimensionality reduction (MDR) featured cross-validation to mitigate over-fitting, computationally efficient searches of the epistatic model space, and variable construction with constructive induction to alleviate the curse of dimensionality. However, the method was unable to differentiate association signals arising from true interactions from those due to independent main effects at individual loci. This issue leads to problems in inference and interpretability for the results from MDR and the family-based compliment the MDR-pedigree disequilibrium test (PDT). A suggestion from previous work was to fit regression models post hoc to specifically evaluate the null hypothesis of no interaction for MDR or MDR-PDT models. We demonstrate with simulation that fitting a regression model on the same data as that analyzed by MDR or MDR-PDT is not a valid test of interaction. This is likely to be true for any other procedure that searches for models, and then performs an uncorrected test for interaction. We also show with simulation that when strong main effects are present and the null hypothesis of no interaction is true, that MDR and MDR-PDT reject at far greater than the nominal rate. We also provide a valid regression-based permutation test procedure that specifically tests the null hypothesis of no interaction, and does not reject the null when only main effects are present. The regression-based permutation test implemented here conducts a valid test of interaction after a search for multilocus models, and can be applied to any method that conducts a search to find a multilocus model representing an interaction.

AB - The initial presentation of multifactor dimensionality reduction (MDR) featured cross-validation to mitigate over-fitting, computationally efficient searches of the epistatic model space, and variable construction with constructive induction to alleviate the curse of dimensionality. However, the method was unable to differentiate association signals arising from true interactions from those due to independent main effects at individual loci. This issue leads to problems in inference and interpretability for the results from MDR and the family-based compliment the MDR-pedigree disequilibrium test (PDT). A suggestion from previous work was to fit regression models post hoc to specifically evaluate the null hypothesis of no interaction for MDR or MDR-PDT models. We demonstrate with simulation that fitting a regression model on the same data as that analyzed by MDR or MDR-PDT is not a valid test of interaction. This is likely to be true for any other procedure that searches for models, and then performs an uncorrected test for interaction. We also show with simulation that when strong main effects are present and the null hypothesis of no interaction is true, that MDR and MDR-PDT reject at far greater than the nominal rate. We also provide a valid regression-based permutation test procedure that specifically tests the null hypothesis of no interaction, and does not reject the null when only main effects are present. The regression-based permutation test implemented here conducts a valid test of interaction after a search for multilocus models, and can be applied to any method that conducts a search to find a multilocus model representing an interaction.

UR - http://www.scopus.com/inward/record.url?scp=77949617596&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77949617596&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0009363

DO - 10.1371/journal.pone.0009363

M3 - Article

C2 - 20186329

AN - SCOPUS:77949617596

VL - 5

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 2

M1 - e9363

ER -