Transpose a large dataset

Question

[ayanabha.research01]

I have a gene expression dataset in Excel/CSV file with 200 columns and 22,000 rows. What will be the best way to switch columns and rows. (This cannot be done in Excel as it supports maximum ~ 16,000 columns)

I want my output as CSV file so that I can load it in Weka.

I don't know programming...So it would be good to know if there is any available tool for this

Answer 1

[Nicolas Rosewick]

Never use excel for bioinformatics stuff!

Use R.

In R:

a <- read.table(file.csv,sep="\t")
b <- t(a)

[ayanabha.research01]

Hi NicoBxl... I tried this with a small dataset , it seem to work fine...but I guess there are some issues working with such large dataset ..it gives some error

[Nicolas Rosewick]

200 x 22,000 should work perfectly ;) What is the amount of RAM you have?

[ayanabha.research01]

My laptop has 4GB RAM

[Nicolas Rosewick]

should work

[ayanava18]

Tried..

Giving output like this...confused

[,22239]                      [,22240]                  [,22241]                 
    [,22242]                  [,22243]         [,22244]         [,22245]        
    [,22246]           [,22247]           [,22248]           [,22249]        
    [,22250]         [,22251]         [,22252]              [,22253]             
    [,22254]              [,22255]              [,22256]             
    [,22257]              [,22258]              [,22259]             
    [,22260]              [,22261]                [,22262]               
    [,22263]                [,22264]               [,22265]              
    [,22266]               [,22267]               [,22268]              
    [,22269]               [,22270]               [,22271]                  
    [,22272]                 [,22273]                   [,22274]                
    [,22275]                 [,22276]                 [,22277]             
    [,22278]              [,22279]         [,22280]         [,22281]        
    [,22282]          [,22283]          [,22284]         
 [ reached getOption("max.print") -- omitted 11 rows ]

[swbarnes2]

Note one little quirk of read.table, it expects the header row to have one less entry than the rest of the file, so if you have a header for the first column, it's going to behave strangely. Also note that you might need to manually add a space to the header after you use write.table, or the header names might not match columns correctly.

[ayanabha.research01]

Also how do I export the transposed data from R to my desktop as a CSV file?

[Nicolas Rosewick]

write.table(b,"out.txt",sep="\t",quote=F)

[Irsan]

write.csv(b,"out.csv")

[ayanava18]

Tried this code, but its exporting the original file but not the transposed one

[Edalat]

Dear NicoBxl,

how to change in the structure and the data types of the data frame after the transposition?

[Nicolas Rosewick]

Better rewrite a new thread for your question in order to be more specific.

Answer 2

[5heikki]

Transpose is handy GNU core util style program..