Hi,
I am looking to analysis a microarray dataset, generated by another lab which I have obtained via getGEO
. Since this is my first time analyzing a microarray dataset, I have a difficult time understanding whether the data I have downloaded represents normalized gene expression or results from differential expression analyses that the author had performed beforehand (hence I should not use these values for another DE analysis).
The data was generated from the platform GPL6480
and the description of the raw data is log2 ratio of test/reference
. From the methods section, samples were labelled with CY5 (test channel)
, and hybridized with CY3 reference channel pooled by normal samples
to generate a log ratio. So my question is, whether I treat these as gene expression values and perform a subsequent differential expression analysis between my samples of interest? Or are these inherently the DE results itself, therefore I cannot use it to compare between samples?
It seems that this is two-color microarray data, so you can perform a DE analysis using methods designed for two-color microarrays. But it would clarify matters if you explain which dataset you have downloaded, i.e., give the GSE dataset number. Then it would be possible to give specific advice. Giving the GPL number only explains the probe set, it doesn't explain the meaning of the channels or how the data have been preprocessed.