(1)生存分析的KM曲线绘制问题
在绘制之前，我们会

group <- ifelse(gene > = median(gene), 'high', 'low')

一般常用中位数将样本分为高低表达组，这样便于绘制，但是假如说某个基因表达量为0的样本数目超过了半数，这样的结果就是所有该基因的所有表达量被修改成‘high’，这样会导致，生存曲线绘制错误。

error in ggsurvplot_df(d, fun = fun, color = color, palette = palette, :
The length of legend.labs should be 1

修改的代码是

group <- ifelse(gene > median(gene), 'high', 'low') 取消等号

==============================================

对表达量矩阵进行反log2+1

a <- data.frame(a = c(1, 1, 5, 6, 8, 9),
                b = c(2, 7, 5, 2, 6, 9),
                c = c(1, 4, 9, 0, 8, 4))


a_new <-   data.frame(matrix(nrow=nrow(a),ncol=ncol(a)))


for (i in 1:nrow(a)) {
  for (j in 1:ncol(a)) {
   
    num <- a[i,j]
    a_new[i,j] <- 2^num - 1
    
  }
  
}

rownames(a_new) <- rownames(a)
colnames(a_new) <- colnames(a)

a
a_new