Related Experiment Video
Updated: Jan 5, 2026
13:19
Microarray-based Identification of Individual HERV Loci Expression: Application to Biomarker Discovery in Prostate Cancer
Published on: November 2, 2013
17.0K
Random forest-based modelling to detect biomarkers for prostate cancer progression
Reka Toth1, Heiko Schiffmann2, Claudia Hube-Magg2
1Cancer Epigenomics, German Cancer Research Center (DKFZ), 69120, Heidelberg, Germany.
Clinical Epigenetics
|October 24, 2019
Summary
This study developed a DNA methylation classifier to predict aggressive prostate cancer (PCa). The model accurately identifies patients with poor prognosis, potentially reducing overtreatment and improving clinical management.
Area of Science:
- Genomics
- Cancer Biology
- Bioinformatics
Background:
- Prostate cancer (PCa) exhibits variable clinical behavior, necessitating personalized treatment strategies.
- Overtreatment of indolent PCa can lead to adverse effects and increased healthcare costs.
- Robust prognostic markers are crucial for guiding treatment decisions and avoiding unnecessary interventions.
Purpose of the Study:
- To develop and validate a random forest-based classification model predicting aggressive prostate cancer behavior using DNA methylation data.
- To identify novel DNA methylation markers and genes associated with PCa progression.
- To improve clinical management of PCa by providing accurate prognostic information.
Main Methods:
- Genome-wide DNA methylation analysis using Illumina HumanMethylation450 arrays on PCa tissues.
- Development of a random forest classification model trained on a discovery cohort (n=70).
- External validation of the model using ICGC (n=222) and TCGA PRAD (n=477) cohorts.
- Immunohistochemistry validation of candidate genes (e.g., ZIC2) in over 12,000 PCa cases.
Main Results:
- The methylation-based classifier achieved high accuracy in predicting PCa prognosis (AUC=95% in test set).
- External validation demonstrated good performance in independent cohorts (AUCs of 77.1% and 68.7%).
- Loss of ZIC2 protein expression was significantly associated with poor prognosis and shorter time to biochemical recurrence, independent of established variables.
Conclusions:
- DNA methylation changes, particularly hypomethylation in partially methylated domains (PMDs), are relevant prognostic indicators for PCa.
- The developed classification model, along with protein expression analysis of identified genes like ZIC2, can aid in clinical decision-making for PCa management.
- This approach supports personalized therapy by distinguishing aggressive from indolent prostate cancer cases.
