12、MA图的计算过程

为了简化问题，假设有3张芯片，每组数有9个探针：

 

Data：

2,4,6,7,9,10,4,7,8,3

9,5,3,2,5,7,9,10,3,12

6,4,3,2,7,8,1,2,6,9

 

 

一、给3组数取2的对数

 

Log2Data：

1.0,2.0,2.5849626,2.807355,3.169925,3.321928,2.0,2.807355,3.0

3.169925,2.321928,1.5849625,1.0,2.321928,2.807355,3.169925,3.321928,1.5849625

2.5849626,2.0,1.5849625,1.0,2.807355,3.0,0.0,1.0,2.5849626

 

二、上面的结果有9列，给每一列取中位数：

 

Median：

2.5849626,2.0,1.5849625,1.0,2.807355,3.0,2.0,2.807355,2.5849626

 

四、M的值为每一列减去该列中位数：

 

M = Log2Data-Median：

-1.5849625,0.0,1.0,1.8073549,0.36257008,0.32192808,0.0,0.0,0.4150375

0.5849625,0.32192808,0.0,0.0,-0.4854268,-0.19264507,1.169925,0.51457316,-1.0

0.0,0.0,0.0,0.0,0.0,0.0,-2.0,-1.8073549,0.0

 

五、A的值是每一列的值加上该列中位数后*0.5

 

A = 0.5*(data_log+medianchip)

1.7924813,2.0,2.0849626,1.9036775,2.98864,3.160964,2.0,2.807355,2.7924812

2.8774438,2.160964,1.5849625,1.0,2.5646415,2.9036775,2.5849626,3.0646415,2.0849626

2.5849626,2.0,1.5849625,1.0,2.807355,3.0,1.0,1.9036775,2.5849626

 

六、以上的M、A值每一行代表一个样品，以M为y轴坐标，A为x轴坐标，绘制出来的散点图就是MA图

 

七．每个样品的Sigma=每个样品的M的中位数

 

0.32192808

0

0

 

八、每个样品的IQR=每个样品的M的IQR

 

0.4150375

0.7072182

0