Center for Clinical and Translational Sciences, University of Texas Health Science Center at Houston, Houston, Texas 77030, USA.
Bioinformatics. 2012 Apr 15;28(8):1109-13. doi: 10.1093/bioinformatics/bts093. Epub 2012 Feb 24.
As an epigenetic alteration, DNA methylation plays an important role in epigenetic controls of gene transcription. Recent advances in genome-wide scan of DNA methylation provide great opportunities in studying the impact of DNA methylation on many human diseases including various types of cancer. Due to the unique feature of this type of data, applicable statistical methods are limited and new sophisticated approaches are desirable.
In this article, we propose a new statistical test to detect differentially methylated loci for case control methylation data generated by Illumina arrays. This new method utilizes the important finding that DNA methylation is highly correlated with age. The proposed method estimates the overall P-value by combining the P-values from independent individual tests each for one age group. Through real data application and simulation study, we show that the proposed test is robust and usually more powerful than other methods.
作为一种表观遗传改变，DNA 甲基化在基因转录的表观遗传调控中起着重要作用。全基因组 DNA 甲基化扫描的最新进展为研究 DNA 甲基化对包括各种类型癌症在内的许多人类疾病的影响提供了极好的机会。由于这类数据的独特特征，适用的统计方法有限，因此需要新的复杂方法。
在本文中，我们为 Illumina 微阵列生成的病例对照甲基化数据提出了一种新的统计检验方法，用于检测差异甲基化基因座。该新方法利用了 DNA 甲基化与年龄高度相关的重要发现。所提出的方法通过组合每个年龄组的独立个体检验的 P 值来估计总体 P 值。通过真实数据应用和模拟研究，我们表明所提出的检验是稳健的，通常比其他方法更有效。
