gms | German Medical Science

49. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds)
19. Jahrestagung der Schweizerischen Gesellschaft für Medizinische Informatik (SGMI)
Jahrestagung 2004 des Arbeitskreises Medizinische Informatik (ÖAKMI)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie
Schweizerische Gesellschaft für Medizinische Informatik (SGMI)

26. bis 30.09.2004, Innsbruck/Tirol

Statistical Analysis of Gen Methylation Data

Meeting Abstract (gmds2004)

  • corresponding author presenting/speaker Georg Göbel - Medical University of Innsbruck, Innsbruck, Österreich
  • Hannes M. Mueller - Medical University of Innsbruck, Innsbruck, Österreich
  • Heidi Fiegl - Medical University of Innsbruck, Innsbruck, Österreich
  • Martin Widschwendter - Medical University of Innsbruck, Innsbruck, Österreich
  • Karl-Peter Pfeiffer - Medical University of Innsbruck, Innsbruck, Österreich

Kooperative Versorgung - Vernetzte Forschung - Ubiquitäre Information. 49. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds), 19. Jahrestagung der Schweizerischen Gesellschaft für Medizinische Informatik (SGMI) und Jahrestagung 2004 des Arbeitskreises Medizinische Informatik (ÖAKMI) der Österreichischen Computer Gesellschaft (OCG) und der Österreichischen Gesellschaft für Biomedizinische Technik (ÖGBMT). Innsbruck, 26.-30.09.2004. Düsseldorf, Köln: German Medical Science; 2004. Doc04gmds084

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/meetings/gmds2004/04gmds084.shtml

Published: September 14, 2004

© 2004 Göbel et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Outline

Text

Introduction

DNA methylation is a common molecular alteration in colorectal cancer (CRC) cells. Recently we report the first time about analysis of fecal DNA from patients with CRC to determine the feasibility, sensitivity and specificity of this approach [1]. Using MethyLight analysis in fecal DNA of three independent sets of patients, we identified SFRP2 methylation to represent a sensitive single DNA-based marker for identification of CRC in stool samples (sensitivity of 90% [CI 56%; 100%] and specificity of 77% [CI 46%; 95%] in the training set and sensitivity of 77% [CI 46%; 95%] and specificity of 77% [CI 46%; 95%] in an independent test set). In this article we want to present the study design and several experiences with Gen Methylation Data.

Methods

This proof of principle study aimed to clarify whether it is possible to use methylation changes in fecal DNA isolated from stool samples as a screening tool for CRC. We designed a three-step, prospective study aiming on the one hand to evaluate the most promising epigenetic markers for CRC out of a long list of candidate genes (gene evaluation set) and, on the other hand, to eventually test these genes in two independent sets of patients (training and test set).

Results

During the first step we used three different statistical methods for the selection of potential tumor marking genes: Mann-Whitney U test by using PMR (Percentage of fully Methylated Reference) values, chi-squared contingency test and PAM (Prediction Analysis for Microarrays) using shrunken centroid clustering method [2]. This resulted in a list of ten promising genes. Assessing the genes in independent training and test sets SFRP2 was identified to be highly methylated in fecal DNA of cancer patients, whereas most of control patients did not show methylation of SFRP2. With this proof of principle study we show that a single DNA-based marker (SFRP2), analysed in independent sets of patients, yields a sensitivity of 77% (test set) to 90% (training set) and a specifity of 77% (both sets) for identifying patients with CRC / non CRC.

Discussion and Outlook

From the statistical viewpoint several issues must be considered: Data structure and statistical distributions of gene methylation data (PMR values) are an upcoming topic in bioinformatics. Advanced methods like bootstrapping or cernel estimation may help researchers to design more sophisticated experimental designs. The statistical power of the combination of different methods for gene selection must be evaluated by further studies. Based on the very promising results the comprehensive development of high specific screening test for CRC will be a challenging task for the future.


References

1.
Mueller HM et al. Methylation changes in faecal DNA - a tool for colorectal cancer screening? Lancet - accepted for publication on 25th March.
2.
Tibshirani R, Hastie T, Narasimhan B, Chu G. Diagnosis of multiple cancer types by shrunken centroids of gene expression. Proc Natl Acad Sci U S A. 2002 May 14;99(10):6567-72.
3.
Mueller HM, Widschwendter A, Fiegl H, Ivarsson L, Goebel G, Perkmann E, Marth C, Widschwendter M. DNA methylation in serum of breast cancer patients: an independent prognostic marker. Cancer Res 2003; 63: 7641-5.