This is a test version of Biostars. For the public version, visit https://www.biostars.org.
R programming: genotype concordance

Hi guys,

I have an R programming question. I want to compare the genotype (.GT columns) with the given alleles (.allele columns) in dataframe df1 and see if they are concordant or not. The rule is that if there is only one allele, it should be 0/0. If there are two alleles (for example GA), the genotype should be 0/1 and that is why I have mismatch in the concordance column for GA. So, A, T, G or C individually is 0/0 and in pair combination is 0/1. Based on this information, I want to add new concordance column next to every two columns that are compared with and have match or mismatch result in concordance column. This concordance column should be cbind next to every compared pair of column and get Result. Could you please help me get this done. Thank you.

df1

1.allele  1.GT  2.allele  2.GT
A         0/0   A         0/0
GA        0/0   CT        0/1
C         0/0   G         0/0

Result

1.allele  1.GT  1.Concordance  2.allele  2.GT  2.Concordance
A         0/0   match          A         0/0   match
GA        0/0   mismatch       CT        0/1   match
C         0/0   match          G         0/0   match
r
• 5,307 views • link
updated by Ram Ram • written by MAPK MAPK

What is the second "Concordance" column for?

• 0 views • link
written by Sean Davis Sean Davis

second concordance is for 2. columns, and since there is no mismatch it's left blank.

• 0 views • link
written by MAPK MAPK

Are there also homozygous SNPs, such as GG or TT?

• 0 views • link
written by Jimbou Jimbou

No, there are not, Thank you.

• 0 views • link
written by MAPK MAPK

2 answers

Data:

df <- data.frame(allele1=c("A","AT","C"),GT1=c("0/0","0/0","0/0"),
allele2=c("AT","G","CG"),GT2=c("0/0","0/0","0/0"))

First I create a vector of correct translation. But I am assuming that there is always a 0/1, never a 1/0:

library(stringr)
translation <- function(x) ifelse(str_length(df[,x])>1,"0/1","0/0")
correct <- sapply(c(1,3),translation)

Finally, mis/match:

match.fun <- function(x) ifelse(df[,x] == correct[,(x/2)] , "match" , "mismatch"  )
comparison <- sapply(c(2,4),match.fun)
cbind(df[,1:2],"Concordance1"=comparison[,1],df[,3:4],"Concordance2"=comparison[,2])
• 0 views • link
updated by Ram Ram • written by Jimbou Jimbou

This is awesome, Thank you!

• 0 views • link
written by MAPK MAPK

The following should do what you need, assuming that your data.frame is called dat.

dat$concordance = ifelse(dat[,2]==dat[,5],"","mismatch")
• 0 views • link
updated by Ram Ram • written by Sean Davis Sean Davis

Sorry, looks like you did not understand the question. Could you please check the updated question. Thanks

• 0 views • link
written by MAPK MAPK

Log in to answer this question.

More posts like this