Fisher exact test in R

Question

I have data as shown below:

chr1    876499  A   517    G    948 A   353  G  964
chr1    877715  C   1166   G    883 C   555  G  944
chr1    877831  T   0      C    1   T   0    C  1
chr1    878314  G   933    C    666 G   89   C  106
chr1    878331  C   983    T    166 C   994  T  505

I would like to do fisher exact test, by making a matrix from columns 4,6,8 and 10 in the below shown way:

For example for first row:

       u    v
X   517   948
Y   353   964

It should calculate the Fisher-exact p-value, and add it as a new column to each row. Since im novice to R i do not have any code. Any help is valuable.

Thanks.

Have you tried anything yet, or searched for examples of how to perform this test in R, or typed ?fisher.test into your R console? — Steve Kern
yes, but haven't been successful in making the matrix as shown. — chas
@rawr Thanks. could you let me know assuming count data in the above data frame? — chas

Andri Signorell Andri Signorell · Accepted Answer · 2015-03-04T22:54:01

Try:

 d.frm <- read.table(textConnection("
 chr1    876499  A   517    G    948 A   353  G  964
 chr1    877715  C   1166   G    883 C   555  G  944
 chr1    877831  T   0      C    1   T   0    C  1
 chr1    878314  G   933    C    666 G   89   C  106
 chr1    878331  C   983    T    166 C   994  T  505
 "))

d.frm$pval <- apply(as.matrix(d.frm[, c(4,6,8,10)]), 1, 
                    function(x) fisher.test(matrix(x, nrow=2))$p.value)

Fisher exact test in R

1 Answers