如何给R语言DataFrame添加基于三条件的threshold列

现有如下R DataFrame：

dat <- data.frame(
  geneID = c("ENSG00000000003.14", "ENSG00000000419.12", "ENSG00000000457.13", "ENSG00000000460.16",
             "ENSG00000001036.13", "ENSG00000001084.10", "ENSG00000001167.14", "ENSG00000001460.17"),
  baseMean = c(2700.791337, 1571.143316, 526.2282051, 1108.138705, 2662.132047, 1325.447272, 1829.828657, 641.7582879),
  log2FoldChange = c(-0.345466785, -0.348258736, -0.051250213, -0.078538637, 0.121419414, 0.89, -0.221749678, -0.252419377),
  lfcSE = c(0.202389477, 0.150807514, 0.180482116, 0.167859597, 0.175209898, 0.154875429, 0.153100403, 0.183602552),
  stat = c(-1.706940451, -2.309293001, -0.283962835, -0.467882913, 0.692994033, -0.423289781, -1.448393819, -1.374814095),
  pvalue = c(0.087833121, 0.020927328, 0.776438862, 0.639868323, 0.488313296, 0.672083849, 0.147506943, 0.169189087),
  padj = c(0.001, 0.120478416, 0.003, 0.827329552, 0.728842774, 0.0004, 0.386446872, 0.417816879)
)

需要添加名为threshold的列，满足以下规则：

• 当log2FoldChange > 0且padj < 0.05时，标记为up
• 当log2FoldChange < 0且padj < 0.05时，标记为down
• 其他情况标记为NS

尝试了以下代码但未达到预期效果：

dat <- mutate(dat, threshold=if_else(dat$padj <= 0.05 & dat$log2FoldChange > 0,"up","NS"))
dat <- mutate(dat, threshold=if_else(dat$padj <= 0.05 & dat$log2FoldChange < 0,"down","NS"))

两次mutate操作会覆盖之前的结果：第一次标记的up会在第二次mutate中被重置为NS，因为第二次只处理log2FoldChange < 0的情况，不满足该条件的行都会被设为NS，导致之前的up标记丢失。

方法1：嵌套if_else #

在一次mutate中完成多条件判断，避免结果被覆盖：

library(dplyr)

dat <- dat %>%
  mutate(threshold = if_else(padj < 0.05 & log2FoldChange > 0, "up",
                            if_else(padj < 0.05 & log2FoldChange < 0, "down", "NS")))

方法2：使用case_when（更清晰，适合多条件场景） #

case_when可以按顺序匹配条件，代码可读性更强，适合复杂的多分支判断：

library(dplyr)

dat <- dat %>%
  mutate(threshold = case_when(
    padj < 0.05 & log2FoldChange > 0 ~ "up",
    padj < 0.05 & log2FoldChange < 0 ~ "down",
    TRUE ~ "NS"  # 所有不满足上述条件的情况统一标记为NS
  ))

运行上述代码后，得到的dat与期望输出一致：

print(dat)
#                 geneID  baseMean log2FoldChange      lfcSE       stat    pvalue      padj threshold
# 1 ENSG00000000003.14 2700.7913    -0.345466785 0.20238948 -1.7069405 0.08783312 0.00100000      down
# 2 ENSG00000000419.12 1571.1433    -0.348258736 0.15080751 -2.3092930 0.02092733 0.12047842        NS
# 3 ENSG00000000457.13  526.2282    -0.051250213 0.18048212 -0.2839628 0.77643886 0.00300000      down
# 4 ENSG00000000460.16 1108.1387    -0.078538637 0.16785960 -0.4678829 0.63986832 0.82732955        NS
# 5 ENSG00000001036.13 2662.1320     0.121419414 0.17520990  0.6929940 0.48831330 0.72884277        NS
# 6 ENSG00000001084.10 1325.4473     0.890000000 0.15487543 -0.4232898 0.67208385 0.00040000        up
# 7 ENSG00000001167.14 1829.8287    -0.221749678 0.15310040 -1.4483938 0.14750694 0.38644687        NS
# 8 ENSG00000001460.17  641.7583    -0.252419377 0.18360255 -1.3748141 0.16918909 0.41781688        NS

内容的提问来源于stack exchange，提问作者user3138373