Hello,
I am not only interested in finding out which genes are the most highly up-
or down-regulated (which I have done using the linear models and Bayesian
statistics in Limma), but I also want to know which genes are consistently
highly transcribed (ie. they have a high intensity in the channel of
interest eg. Cy5 or Cy3 across the set of experiments). I might have missed
a straight forward way to do this, or a valuable function, but I've been
using my own methods and going around in circles.
So far I've normalized within and between arrays, then returned the RG
values using RG<-RG.MA, then I ranked each R and G values for each array as
below.
rankRG<-RG
rankRG$R[,1]<-rank(rankRG$R[,1])
rankRG$R[,2]<-rank(rankRG$R[,2]) .. and so on for 6 columns(ie. arrays, as
well as the G's)
then I thought I could pull out a subset of rankRG using something like;
topRG<-rankRG
topRG$R<-subset(topRG$R,topRG$R[,1]<500&topRG$R[,2]<500&topRG$R[,5]<500)
However, this just returned me a matrix with one row of $R (the ranks were
<500 for columns 1,2, and 5 and greater than 500 for 3,4,and 6). However, I
can't believe that there is only one gene that is in the top 500 for R
intensitiy among those three arrays.
Am I doing something wrong? Can someone think of a better way of doing
this?
Thanks
Alison
******************************************
Alison S. Waller M.A.Sc.
Doctoral Candidate
awaller@chem-eng.utoronto.ca
416-978-4222 (lab)
Department of Chemical Engineering
Wallberg Building
200 College st.
Toronto, ON
M5S 3E5
[[alternative HTML version deleted]]