CODEWITHSUNDEEP
r
Cahidora
Cahidora

Reputation: 143

Error when subsetting

I have a problem with my DFs in R. I have a df1 with gene names.

 V1        V2
GENE A     GENE E
GENE B     GENE D
GENE C     GENE A
GENE D     GENE B
GENE E     GENE C

and another dataframe called df2 like this:

Name       ID     Symbol
GENE A    1254    AKT
GENE B    1879    POU5F1
GENE C    5689    EGR1
GENE D    2385    JUN
GENE E    5687    MYC

The output I would like to have is the following:

NameSource       SourceID        NameTarget       TargetID
AKT                1254             MYC             5687
POU5F1             1879             JUN             2385
EGR1               5689             AKT             1254
JUN                2385             POU5F1          1879
MYC                5687             EGR1            5689

I tried with the following syntax:

genes <- df1[which(df1$V1, df2$Symbol), ]

and with:

genes <- df1$V1 %in% df2$Symbol

But for some reason I cannot get the output I am expecting. Anyone?

Upvotes: 2

Views: 56

Answers (2)

Bastien
Bastien

Reputation: 3098

with dplyr and piping if you're in that kind of R:

df1 <-  data.frame(V1=c("GENE A","GENE B","GENE C","GENE D", "GENE E"), V2=c("GENE E","GENE D","GENE A","GENE B", "GENE C"))

df2 <- data.frame(Name= c("GENE A","GENE B","GENE C","GENE D", "GENE E"),
                  ID= c(1254,1879,5689,2385,5687),
                  Symbol= c("AKT", "POU5F1", "EGR1", "JUN", "MYC")
)

library(dplyr)

df1 %>% right_join(df2, by=c("V1"="Name")) %>% 
   right_join(df2, by=c("V2"="Name")) %>% 
   arrange(V1) %>% 
   select(ID.x, Symbol.x, ID.y, Symbol.y) %>% 
   setNames(c("SourceID", "NameSource", "TargetID",  "NameTarget"))

#  SourceID NameSource TargetID NameTarget
#1     1254        AKT     5687        MYC
#2     1879     POU5F1     2385        JUN
#3     5689       EGR1     1254        AKT
#4     2385        JUN     1879     POU5F1
#5     5687        MYC     5689       EGR1

Upvotes: 3

akrun
akrun

Reputation: 887158

We can use match

res <- cbind(df2[-1], df2[match(df1$V2, df2$Name),-1])[c(2,1,4 ,3)]
colnames(res) <- c("NameSource", "SourceID", "NameTarget", "TargetID")
row.names(res) <- NULL
res
#   NameSource SourceID NameTarget TargetID
#1        AKT     1254        MYC     5687
#2     POU5F1     1879        JUN     2385
#3       EGR1     5689        AKT     1254
#4        JUN     2385     POU5F1     1879
#5        MYC     5687       EGR1     5689

Upvotes: 4

Related Questions