2
$\begingroup$

After scanning the microarray chip we get RAW data with different formats depending on the type of microarray used. However, these RAW data files can be normalized and processed to generate gene expression data matrix.

I've been trying to figure out what exactly is the result of the processed data and what is the format of that data. As I've seen so far, there's no specific format for processed data. Of course most of them agree that columns represent experiments and rows represent genes, but the additional information at the top, they're usually different but I'm not sure based on what. Can anyone suggest me with the most common used formats? Is there any additional information to that processed data file in case I wanted to apply classification on it?

$\endgroup$
  • $\begingroup$ Just to be clear, it sounds like you're asking how microarray data is typically displayed, as in a publication. This would be different from the actual content of the processed data. In the simplest case, processed microarray data consists of a numerical value for each point on the chip, i.e. each gene you are testing. For each point on the chip (i.e. for each gene), the data is the ratio of mRNA abundance from one experimental condition vs another. You can imagine a lot of ways this data might be presented, e.g. a heat map of relative abundance of mRNA for each gene. $\endgroup$ – A. Kennard Nov 9 '13 at 17:16
  • $\begingroup$ @A.Kennard Do you mean it depends on the type of microarray used? Affymetrix has different Array design than illumnia for example $\endgroup$ – Lamia Nov 13 '13 at 21:26
  • 1
    $\begingroup$ I think I might not be understanding your question. What do you mean by "the additional information at the top"? Could you provide an example of processed microarray data you are having trouble understanding? $\endgroup$ – A. Kennard Nov 13 '13 at 23:54
2
$\begingroup$

The RAW signal values from each microarray are transformed and processed to give you normalized signal values for each microarray. The transformation and processing steps are standard but different for each manufacture. Each manufacture uses a different method/technology to obtain signal values. Thus, signal values are not directly comparable between different microarray technologies.

Experiments using different technologies might get similar RANK of results but the normalized signal values will be different. You can have a look at GEO for a list of datasets and commonly used technologies. I have no clue what you mean by applying classification (are you looking to cluster the results from different microarray technologies run on the same experiment? I'm pretty sure a comparison of microarray technologies has been done before/already)

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.