[R] Help with correlation matrices, thresholding

chuck.01 CharlieTheBrown77 at gmail.com
Mon Jul 16 23:28:41 CEST 2012

Try this as a starting point; dat is your whole data set, and r = correlation
coefficient threshold:

 spec.cor <- function (dat, r, ...) 
    x <- cor(dat, ...)
    x[upper.tri(x, TRUE)] <- NA
    i <- which(abs(x) >= r, arr.ind = TRUE)
    data.frame(matrix(colnames(x)[as.vector(i)], ncol = 2), value = x[i])

Drinniol wrote
> Hello, I am new to R and have a problem I have had some trouble with.  
> Basically I have a list of some 80000 genes or so with data points for
> expression levels at various time points/conditions.  I also have subsets
> of these, usually only a few hundred genes in size, known to be associated
> with some biological process.
> What I want to do is correlate the entire list with my subset and then
> return out the gene names and correlations for correlations above a
> certain threshold - the idea being that these would be good candidate
> genes to look at more closely for possible interaction with said
> biological process.  
> What functions should I use to correlate the subset to the whole set -
> keeping in mind that I need to keep track of gene names in both sets?  And
> how, then, from what I presume would be something like a 80000 x 200 sized
> matrix would I extract gene pair correlations over a certain threshold? 
> It would also probably be good if I could somehow remove the subset genes
> from the whole gene list before computation so as to avoid useless 1
> correlations.  
> I'll keep trying on my own, but any help would be appreciated!

View this message in context: http://r.789695.n4.nabble.com/Help-with-correlation-matrices-thresholding-tp4636697p4636704.html
Sent from the R help mailing list archive at Nabble.com.

More information about the R-help mailing list