DESeq Count Tables Prep

[coralgirl]

Hi,

I'm new to almost all of this - R, DESeq, bioinformatics etc... (obviously a bench biologist).

I've been given a count table from an alignment run with a custom Perl script that is a tab delimited text file. The header includies the ContigID title, and each sample name, and then each row contains the contig number, and the counts for each sample.

How do I use this as the count file for DESeq? Can I import it as a tab delim. txt file, or do I have to change it into a file that is readable by DESeq?

I just want to be able to get to the point where I can walk through the DESeq vignette on R with my own data.

Thanks

[coralgirl]

An addition to my post:

So I've uploaded the txt file successfully as a read.table, and uploaded library(DESeq) and am now trying to get my file to be the countsTable, and this is where I'm running into problems. My R code:

> countsTable <- counts( myFile )
Error in function (classes, fdef, mtable) :
unable to find an inherited method for function "counts", for signature "data.frame"

How do I make my read.table file (i.e. the uploaded tab delim. txt file) to act as the countsTable?

Thanks

[kopi-o]

Have you tried the newCountDataSet() function? It should work if your file that you imported with read.table() is in the correct format, which it may very well be from your description. You're not supposed to use the counts() function in this case - I think that's just a way to extract counts from a pre-existing CountDataSet object, which you don't have.

[coralgirl]

Ok - so I tried the newCountDataSet() function with the imported read.table() and I received the following error:
Error in round(countData) : Non-numeric argument to mathematical function

This is everything that I've input so far:
> fullmoonDN <- read.table("filelocation",header=TRUE,sep="\t")
> conds <- factor( c("AM0008","AM0009") )
> cds <- newCountDataSet( fullmoonDN, conds )
Error in round(countData) : Non-numeric argument to mathematical function
> #show the first 5 rows
> fullmoonDN[1:5,]
X AM0008 AM0009
1 EZ031385 3045 3218
2 EZ019649 12679 9190
3 EZ010289 800 28050
4 EF202590 816 3108
5 EZ010226 633 459

Is there something wrong with how my txt file is laid out? I did not have any text input for the first column name, and it automatically put an X (I tried with having the first column name as Contig_ID, and I came up with the same result). I'm not sure what I'm doing wrong here. Thoughts?

[kopi-o]

Try to add "row.names=1" to the read.table command. That will make the first column into row names instead of entries in a column. The newCountDataSet() function excepts only numerical values in the input.

[coralgirl]

That worked perfectly. Thank you!