R语言中使用case_when处理双字符列按位匹配生成组合编码的方法

你需要先拆分每列的两个字符，分别对两个位置的碱基组合做规则匹配后再拼接，以下是可直接运行的代码：

首先加载依赖包：

library(dplyr)
library(stringr)

方案1：直接分步处理 #

df_new <- df %>%
  # 提取每列第1位和第2位碱基
  mutate(across(col1:col4, list(pos1 = ~str_sub(.x, 1, 1), pos2 = ~str_sub(.x, 2, 2)), .names = "{.col}_{.fn}")) %>%
  # 对第一个位置的碱基组合匹配规则
  mutate(res1 = case_when(
    col1_pos1 == "A" & col2_pos1 == "C" & col3_pos1 == "T" & col4_pos1 == "A" ~ "01",
    col1_pos1 == "A" & col2_pos1 == "C" & col3_pos1 == "T" & col4_pos1 == "T" ~ "02",
    col1_pos1 == "T" & col2_pos1 == "G" & col3_pos1 == "C" & col4_pos1 == "T" ~ "03",
    col1_pos1 == "T" & col2_pos1 == "G" & col3_pos1 == "C" & col4_pos1 == "A" ~ "04",
    col1_pos1 == "T" & col2_pos1 == "G" & col3_pos1 == "T" & col4_pos1 == "A" ~ "05",
    TRUE ~ "other"
  )) %>%
  # 对第二个位置的碱基组合匹配规则
  mutate(res2 = case_when(
    col1_pos2 == "A" & col2_pos2 == "C" & col3_pos2 == "T" & col4_pos2 == "A" ~ "01",
    col1_pos2 == "A" & col2_pos2 == "C" & col3_pos2 == "T" & col4_pos2 == "T" ~ "02",
    col1_pos2 == "T" & col2_pos2 == "G" & col3_pos2 == "C" & col4_pos2 == "T" ~ "03",
    col1_pos2 == "T" & col2_pos2 == "G" & col3_pos2 == "C" & col4_pos2 == "A" ~ "04",
    col1_pos2 == "T" & col2_pos2 == "G" & col3_pos2 == "T" & col4_pos2 == "A" ~ "05",
    TRUE ~ "other"
  )) %>%
  # 拼接结果，两个都为other时返回other
  mutate(new_col = case_when(
    res1 == "other" & res2 == "other" ~ "other",
    TRUE ~ paste0(res1, "+", res2)
  )) %>%
  # 保留原始列和最终结果列
  select(Id, col1:col4, new_col)

运行后输出的df_new和你给出的预期结果完全一致。

方案2：封装函数简化代码 #

如果要避免重复写两次匹配逻辑，可以把规则封装为自定义函数：

match_rule <- function(c1, c2, c3, c4) {
  case_when(
    c1 == "A" & c2 == "C" & c3 == "T" & c4 == "A" ~ "01",
    c1 == "A" & c2 == "C" & c3 == "T" & c4 == "T" ~ "02",
    c1 == "T" & c2 == "G" & c3 == "C" & c4 == "T" ~ "03",
    c1 == "T" & c2 == "G" & c3 == "C" & c4 == "A" ~ "04",
    c1 == "T" & c2 == "G" & c3 == "T" & c4 == "A" ~ "05",
    TRUE ~ "other"
  )
}

# 调用函数处理
df_new <- df %>%
  mutate(
    res1 = match_rule(str_sub(col1,1,1), str_sub(col2,1,1), str_sub(col3,1,1), str_sub(col4,1,1)),
    res2 = match_rule(str_sub(col1,2,2), str_sub(col2,2,2), str_sub(col3,2,2), str_sub(col4,2,2)),
    new_col = ifelse(res1 == "other" & res2 == "other", "other", paste0(res1, "+", res2))
  ) %>%
  select(Id, col1:col4, new_col)

内容的提问来源于stack exchange，提问作者Marwah Al-kaabi