您可以将数据透视得更宽，并过滤总和仅为1的行：

library(tidyverse)

dat <- tibble(gene = c("abc", "123", "u&me", "acc", "123", "zzz", "efd", "123", "abc"),
       disease = c(rep("DN", 3), rep("FSGS", 3), rep("LN", 3)))

dat %>%
  mutate(val = 1) %>%
  pivot_wider(names_from = disease, values_from = val, values_fill = 0) %>%
  filter(rowSums(.[,2:ncol(.)]) == 1) 
#> # A tibble: 4 x 4
#>   gene     DN  FSGS    LN
#>   <chr> <dbl> <dbl> <dbl>
#> 1 u&me      1     0     0
#> 2 acc       0     1     0
#> 3 zzz       0     1     0
#> 4 efd       0     0     1

字符串
另一种选择是总结独特基因的数量，如果超过1，则进行过滤：

library(tidyverse)

dat |>
  reframe(gene = unique(gene), .by = disease) |>
  mutate(n = n(), .by = gene) |>
  filter(n == 1)
#> # A tibble: 4 x 3
#>   disease gene      n
#>   <chr>   <chr> <int>
#> 1 DN      u&me      1
#> 2 FSGS    acc       1
#> 3 FSGS    zzz       1
#> 4 LN      efd       1

型
第二种方法可能更有效。