How to identify differential methylation sites using GenomeStudio

Today, I will talk to you about how to use GenomeStudio to identify differential methylation sites, which may not be well understood by many people. in order to make you understand better, the editor has summarized the following content for you. I hope you can get something from this article.

Before the difference analysis of methylation chip, we must go through a step of data preprocessing, which includes two steps: normalization and background denoising. Next, let's take a look at the detailed steps of difference analysis in GenomeStudio.

Create a new methylation project

Add the corresponding sample

Determine sample grouping information

Here I randomly selected six samples as control and the other six samples as case.

Set parameters for difference analysis

Analysis Type is set to Diff Methylation

Name represents the name of this analysis.

In particular, normalization and background denoising must be carried out. Normalization choose controls algorithm and check the checkbox in front of Subtract background

Content Descriptor is the comment file in bpm format corresponding to the chip.

Ref Group selected control group

There are three methods for Error Model to choose difference analysis.

Illumina custom

Mann-Whitney

T-test

View the result

By default, there are 3 windows of data

Sample Table

This window contains the basic information of each sample, where Detected CpG represents the probe P value in the sample.

< 0.01 的探针数 Sample Methylation Profile 样本的CpG表达谱, 在这个窗口中包含了CpG位点的表达值，比如Beta值，由于在导入数据时，进行了预处理，所以这里是预处理之后的表达量 Diff methylation 差异甲基化位点的结果 差异分析的结果中，p值时我们最常用的过滤手段，这里默认的结果中并没有给出P值，而是给出了1个DiffScore值。 DiffScore的计算公式如下： DiffScore = 10 sgn(avg_case - avg_control) log10Pvalue Diffscore 其实是和 Pvalue 有联系的，当 P= 0.05 时, DiffScore >

13 or DiffScore

< -13 P = 0.01 时, DiffScore >

22 or DiffScore

< -22 P = 0.001 时, DiffScore >

33 or DiffScore

< -33 所以我们可以根据DiffScore值筛选差异分析的结果，如果你以P=0.05作为阈值， 那么DiffScore >

13 or DIffScore

< -13 的探针就是差异探针了，而且DiffScore >

0 is up-regulated, DiffScore