批量做Log-rank检验

result.KM = data.frame(EnsemblID=NA,p.value=NA)
for(i in 1:304){          #纳入的基因数
  meta$gene = meta[,i]
  meta$group = ifelse(meta$gene > median(meta$gene), 'hign', 'low')
  KM = survdiff(Surv(pfs_time, pfs_status)~group,
              data = meta)
  p.value <- 1 - pchisq(KM$chisq, length(KM$n) - 1) #计算log-rank检验的p值
  
  result.KM[i,1] = names(meta)[i]
  result.KM[i,2] = p.value
} 

• 输入数据 meta 的结构为：基因表达（z-score）+ 生存资料

meta