Removing Duplicate Rows(Gene Names)

Question

[robjohn7000]

Hi,

How can I remove duplicate rows from the output of my toptableF? I have tried the following without success:

  if (unique(fit2$genes)){
  Sigenes.F <- topTableF(fit2, number=100, genelist=fit2$genes, adjust.method="bonferroni",
                   sort.by="F")
  }

Thanks

[Irsan]

I don't think you want to remove duplicated genes, you probably want to keep the best p-value for each gene. Have a look into the aggregate() function

[robjohn7000]

Thanks again, Irsan. I was able to get get aggregate() to work. It makes sense in the way you have explained, and not just throwing genes away.

[brentp]

you can do as Michael Dondrup says below, but that might be removing important information. Often, you have duplicated genes in microarray data because there are multiple probes for a single gene. E.g. what if one of them is up-regulated and one is down-regulated (due to alternative transcript use) and you discard one?

[Joseph Hughes]

Is it from the fit2 data frame or the Sigenes.F data frame that you want to remove the duplicates?

[robjohn7000]

From the Sigenes.F data frame. Thanks.

Answer 1

[Sean Davis]

You'll need to pay attention to what you are doing. If you are new to R, you will need to do some reading on R data structures, as each behaves differently with regard to indexing. Just guessing, but you probably want:

SigGenes.F[!duplicated(SigGenes.F$genes),]

[robjohn7000]

You'r absoultely right. This worked for me as well. Many thanks Sean!

Answer 2

[Michael]

Read ?duplicated:

> x = c(1:10,1:10)
> x
 [1]  1  2  3  4  5  6  7  8  9 10  1  2  3  4  5  6  7  8  9 10
> x[!duplicated(x)]
 [1]  1  2  3  4  5  6  7  8  9 10
>

[robjohn7000]

I used duplicated() this way:

 SigGenes.F[!duplicated(SigGenes.F)]

Got this output:

  Error in `[.data.frame`(SigGenes.F, !duplicated(SigGenes.F)) : 
  undefined columns selected

Then tried:

    SigGenes.F[!duplicated(SigGenes.F[2])]

Output:

      columns 3:7 have been removed

Thanks