[BioC] genereating correlation matrix from gene expression data

[Steve Lianoglou]

Hi Pankaj,

On Nov 18, 2009, at 1:30 PM, pankaj borah wrote:

> Hi all,
>
> i am a new user of R and Bioconductor package. i deal with gene  
> expression data. i was wondering if there is a way to generate a  
> correlation matrix (all to all square symmetric matrix) for  a set  
> of genes and their expression values. is there any available function?

Try ?cor to get the pairwise correlation between columns of your  
matrix, eg:

R> x <- matrix(rnorm(20), 4, 5)
R> x
            [,1]        [,2]       [,3]       [,4]       [,5]
[1,]  0.5818808  0.58781709  2.5511157 -1.5332180  0.6905731
[2,] -0.7640311  0.25960974  0.7246655 -1.5539226 -0.5459625
[3,] -1.7141619 -0.25808091 -0.0868366 -0.6547804  0.4629494
[4,] -2.2906217  0.04932864  0.5694895  0.7736206  0.4078665

R> cor(x)
             [,1]         [,2]       [,3]       [,4]        [,5]
[1,]  1.00000000  0.851982381  0.8715055 -0.8404848 0.074770380
[2,]  0.85198238  1.000000000  0.9429553 -0.5338964 0.004627258
[3,]  0.87150545  0.942955253  1.0000000 -0.4719936 0.325584972
[4,] -0.84048477 -0.533896414 -0.4719936  1.0000000 0.309414781
[5,]  0.07477038  0.004627258  0.3255850  0.3094148 1.000000000

If your genes are in rows, you'll just need to pass in the transpose  
of your matrix.

HTH,

-steve

--
Steve Lianoglou
Graduate Student: Computational Systems Biology
   |  Memorial Sloan-Kettering Cancer Center
   |  Weill Medical College of Cornell University
Contact Info: http://cbio.mskcc.org/~lianos/contact