R语言拆分Intervar空格与等号分隔字符串 方括号内值单独分列

现有R数据集结构如下：

structure(list(Intervar = c("Uncertain significance PVS1=0 PS=[0, 1, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 0, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 1, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Likely benign PVS1=0 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 0, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 1, 0, 0, 1, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]"
)), class = "data.frame", row.names = c(NA, -8L))

Intervar列每一行格式统一：开头为变异临床评级（Pathogenic/Likely pathogenic/Uncertain significance/Likely benign/Benign），后续为多个ACMG证据项赋值，分为两类：

• 单值类型：格式为标识=数值，如PVS1=0
• 多值类型：格式为标识=[数值1,数值2,...]，如PS=[0,1,0,0,0]

需要将Intervar列拆分为多列：

• 单值类型直接生成对应列，列名取等号前的标识
• 多值类型中方括号内每个数值单独成列，列名为标识加序号，如PS对应PS1~PS5

期望输出示例：

PVS1 PS1 PS2 PS3 PS4 PS5 PM1 PM2 PM3 PM4 PM5 PM6 PM7 PP1 PP2 PP3 PP4 PP5 PP6 ... BP1 BP2 BP3 BP4 BP5 BP6 BP7 BP8
1    0   0   1   0   0   0   0   1   0   0   0   0   0   0   0   0   0   0   0 ...   0   0   0   0   0   0   0   0
2    1   0   0   0   0   0   0   1   0   0   0   0   0   0   0   1   0   0   0 ...   0   0   0   0   0   0   0   0
...

此前实现代码会将方括号内数值求和后生成列，仅保留PS、PM等求和后的列，不符合需求：

test$Intervar %>%
  gsub(" (\\w+=)", "\n\\1", .) %>%
  paste0("\nDescription=", .) %>%
  gsubfn("\\[(.*?)\\]", ~ sum(scan(text = txt, sep = ",", quiet = TRUE)), .) %>%
  gsub("=", ": ", .) %>%
  textConnection %>%
  read.dcf %>%
  as.data.frame %>%
  type.convert

以下是基于原有逻辑修改的版本，仅替换了方括号内的处理逻辑，将求和改为生成带序号的键值对：

library(gsubfn)
library(magrittr)
library(stringr)

# 读取原始数据集
test <- structure(list(Intervar = c("Uncertain significance PVS1=0 PS=[0, 1, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 0, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 1, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Likely benign PVS1=0 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 0, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 1, 0, 0, 1, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]", 
"Pathogenic PVS1=1 PS=[0, 0, 0, 0, 0] PM=[0, 1, 0, 0, 0, 0, 0] PP=[0, 0, 1, 0, 0, 0] BA1=0 BS=[0, 0, 0, 0, 0] BP=[0, 0, 0, 0, 0, 0, 0, 0]"
)), class = "data.frame", row.names = c(NA, -8L))

# 核心处理
result <- test$Intervar %>%
  # 移除开头的临床评级
  str_remove("^(Pathogenic|Likely pathogenic|Uncertain significance|Likely benign|Benign) ") %>%
  # 将多值项的[xxx]替换为带序号的键值对，如PS=[0,1,0]转为PS1=0 PS2=1 PS3=0
  gsubfn("(\\w+)=\\[(.*?)\\]", function(key, val_str) {
    vals <- scan(text = val_str, sep = ",", quiet = TRUE)
    paste(paste0(key, seq_along(vals), "=", vals), collapse = " ")
  }, .) %>%
  # 原有格式转换逻辑
  gsub(" (\\w+=)", "\n\\1", .) %>%
  paste0("\n", .) %>%
  textConnection() %>%
  read.dcf() %>%
  as.data.frame() %>%
  type.convert(as.is = TRUE) %>%
  # 按列名排序，可选
  select(order(colnames(.)))

输出的result完全符合需求，单值项直接保留原标识作为列名，多值项自动生成带序号的列。

内容的提问来源于stack exchange，提问作者Lin Caijin