gms | German Medical Science

GMDS 2015: 60. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie

06.09. - 09.09.2015, Krefeld

High-throughput DNA methylation association analyses with reference-free cell type adjustment: A method comparison

Meeting Abstract

  • Miriam Kesselmeier - Klinische Epidemiologie, Integriertes Forschungs- und Behandlungszentrum (IFB) Sepsis und Sepsisfolgen (CSCC), Universitätsklinikum Jena, Jena, Deutschland
  • Anke Hinney - Klinik für Psychiatrie, Psychosomatik und Psychotherapie des Kindes- und Jugendalters LVR-Klinikum Essen, Universitätsklinikum Essen, Universität Duisburg-Essen, Essen, Deutschland
  • Johannes Hebebrand - Klinik für Psychiatrie, Psychosomatik und Psychotherapie des Kindes- und Jugendalters LVR-Klinikum Essen, Universitätsklinikum Essen, Universität Duisburg-Essen, Essen, Deutschland
  • André Scherag - Klinische Epidemiologie, Integriertes Forschungs- und Behandlungszentrum (IFB) Sepsis und Sepsisfolgen (CSCC), Universitätsklinikum Jena, Jena, Deutschland

GMDS 2015. 60. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e.V. (GMDS). Krefeld, 06.-09.09.2015. Düsseldorf: German Medical Science GMS Publishing House; 2015. DocAbstr. 104

doi: 10.3205/15gmds148, urn:nbn:de:0183-15gmds1481

Veröffentlicht: 27. August 2015

© 2015 Kesselmeier et al.
Dieser Artikel ist ein Open-Access-Artikel und steht unter den Lizenzbedingungen der Creative Commons Attribution 4.0 License (Namensnennung). Lizenz-Angaben siehe http://creativecommons.org/licenses/by/4.0/.


Gliederung

Text

Introduction: DNA methylation is one frequently investigated epigenetic mechanism which is cell type dependent. Consequently, small perturbations in cell mixture proportions in whole blood from a case–control study may be registered as changes in DNA methylation [1]. Recently, Houseman et al. [2] and Zou et al. [3] proposed two methods to address cell type admixture bias effects in high-throughput DNA methylation association analyses. In contrast to previous methods, both methods are reference-free, i.e. they do not require a reference data set in which both cell composition and DNA methylation is measured.

Methods: After presenting the two methods using a unified notation, we perform a comparison by a simulation study following the strategy described in [2]. In particular, we contrast empirical error rates in simulated data sets with and without cell type dependency. Furthermore, we vary the sample size and the number of different cell types. Finally, we apply both methods to a real data example of a case-control study.

Results: Under the original simulation scenarios, we observed similar statistical properties of both methods. In the real data example, the cell type composition of cases and controls indeed differed when using the method of [1] which requires a reference. However, we observed an inflation of too small p-values for the reference-free method of Houseman et al. [2] while the reference-free method by Zou et al. [3] seemed to provide better overall type I error control.

Discussion: We discuss the limited overlap of our results in the real data application and conclude by providing some guidance to conduct more realistic simulation studies.


References

1.
Houseman EA, Accomando WP, Koestler DC, Christensen BC, Marsit CJ, Nelson HH et al. DNA methylation arrays as surrogate measures of cell mixture distribution. BMC Bioinform. 2012 May 8; 13:86
2.
Houseman EA, Molitor J, Marsit CJ. Reference-free cell mixture adjustments in analysis of DNA methylation data. Bioinformatics. 2014 May 15; 30(10):1431-9.
3.
Zou J, Lippert C, Heckerman D, Aryee M, Listgarten J. Epigenome-wide association studies without the need for cell-type composition. Nat Methods. 2014 Mar; 11(3):309-11.