Question

我有一个数据框df1：

df <- structure(list(Id = c(0, 1, 3, 4), Support = c(17, 15, 10, 18
), Genes = structure(c(3L, 1L, 4L, 2L), .Label = c("BMP2,TGFB1,BMP3,MAPK12,GDF11,MAPK13,CITED1", 
"CBLC,TGFA,MAPK12,YWHAE,YWHAQ,MAPK13,SPRY4", "FOS,BCL2,PIK3CD,NFKBIA,TNFRSF10B", 
"MAPK12,YWHAE,YWHAQ,MAPK13,SPRY4,PIK3CD"), class = "factor")), class = "data.frame", row.names = c(NA, 
-4L))

和另一个数据框df2：

df2 <- structure(list(V1 = structure(c(6L, 7L, 8L, 4L, 3L, 1L, 5L, 2L, 
9L), .Label = c("BCL2", "BMP3", "CBLC", "CDC23", "CITED1", "FOS", 
"MAPK13", "SPRY4", "TGFA"), class = "factor")), class = "data.frame", row.names = c(NA, 
-9L))

如何通过计算df1列中df2的每个字符串的出现来在Genes中创建新列，以实现所需的输出？

    Id    |    Support    |     Genes    |    Counts    |
---------------------------------------------------------
    0     |       17      |FOS,BCL2,...  |      2       |
    1     |       15      |BMP2,TFGB1,...|      3       |
    3     |       10      |MAPK12,YWHAE..|      1       |
    4     |       18      |CBLC,TGFA,... |      4       |

Answer 1

可能有一个更优雅的解决方案，但这可以完成工作。

df$Counts <- unlist(lapply(df$Genes, function(x){
  xx <- unlist(strsplit(as.character(x),split = ","))
  sum(df2$V1 %in% xx)
}))

哪个给：

 Id Support                                      Genes Counts
1  0      17           FOS,BCL2,PIK3CD,NFKBIA,TNFRSF10B      2
2  1      15 BMP2,TGFB1,BMP3,MAPK12,GDF11,MAPK13,CITED1      3
3  3      10     MAPK12,YWHAE,YWHAQ,MAPK13,SPRY4,PIK3CD      2
4  4      18  CBLC,TGFA,MAPK12,YWHAE,YWHAQ,MAPK13,SPRY4      4

（我认为在您的示例中，第三行Counts上方的应该是2而不是1？）

Answer 2

这是使用纵梁库的另一种选择。这会循环遍历df的Genes列，并使用df2数据帧作为模式。

#convert factors columns into characters
df$Genes<-as.character(df$Genes)
df2$V1<-as.character(df2$V1)

library(stringr)
#loop over the strings against the pattern from df2
df$Counts<-sapply(df$Genes, function(x){
  sum(str_count(x, df2$V1))
})



df
  Id Support                                      Genes Counts
1  0      17           FOS,BCL2,PIK3CD,NFKBIA,TNFRSF10B      2
2  1      15 BMP2,TGFB1,BMP3,MAPK12,GDF11,MAPK13,CITED1      3
3  3      10     MAPK12,YWHAE,YWHAQ,MAPK13,SPRY4,PIK3CD      2
4  4      18  CBLC,TGFA,MAPK12,YWHAE,YWHAQ,MAPK13,SPRY4      4

R：通过计算CSV列中字符串的出现，将计数出现列添加到数据框

2 个答案: