Which Algorithm Should I Choose When I Handle The .Cel When I Download The .Cel From Broad Cmap

Question

[zerohsy]

Hello,when I download the .cel from the Broad CMAP( http://www.broadinstitute.org/cmap/cel_file_chunks.jsp). When I open .cel,what I see is just digital gibberish.So I download the software called Gene Expression from the Aff company.However,there are three algorithm named RMA,MAS5,PLIER WorkFlow to choose and the results are different when choosing different algorithm.So which algorithm should I choose ?And what difference are the three algorithm?

Answer 1

[Obi Griffith]

There is no one best method. If given these three choices I would probably choose RMA or PLIER and avoid MAS5. But, even better would be to use GCRMA or dChip. Describing the differences between these methods would be repeating a vast amount of existing discussion on the topic.

https://www.google.com/search?q=rma+vs+mas5+vs+plier

In particular, you might check out 'Methods in Microarray Normalization' and this paper from Seo and Hoffman for good discussions on the topic. My personal strategy has been to use both standard Affy CDFs and custom CDFs together with GCRMA normalization. Relevant biostar tutorials on the topic:

• Analysing microarray data in BioConductor
• Machine learning for cancer classification - part 1 - preparing the data sets