I would like to analyse microarray samples

Question

[Sabine]

I have two samples with me. they are CEL files. image file form microarray studies. I would like to analyse and determine the differentially expressed genes . Now my question is anybody please tel me how can I analyse the microarray CEL files?

And is there any free software available for this kind of analysis?

Thank you

[Michael]

I can assure you that this is not the right way to get help in any online community. As a rule of thumb, if your question consists of a single sentence you are doing it wrong.

You need to provide much more detail about your data and protocols, and most importantly you need to make an effort yourself and try to define your experimental question. I am not yet talking about choice of software or method at all.

Please read: http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fjournal.pcbi.1002202

[Michael]

I am not aware of Illumina producing cel files, they might be affymetrix cel files, are you sure about the files origin? Do you know anything about the experimental conditions in the study, how many replicates do you have?

[Sabine]

Yes this is from affymetrix, total I have 2 files

[Sabine]

Really? How can I analyse these 2 and give at least a ranking by determining the top genes. Thanks

[Sabine]

And is there any free software available for this kind of analysis?

[Michael]

Please do not make an answer use the add comment button. There is ample free software out there, there are also ample lists of free microarray software out there, here is one: http://pathology.wustl.edu/research/cores/lcg/index.php?page=Microarray%20Data%20Analysis

I recommend R/Bioconductor or dChip

[Sabine]

Thank you..let me try

[Sabine]

I have prepared normalised data as excel using Bioconductor. How can I determine the possible list of genes from that?

I have used 2 methods for the normalization, namely RMA and mas5

Answer 1

[Michael]

If you have only two samples without replication it is not possible to perform a proper DE analysis using statistical tests. The only thing you can do is rank the genes by absolute differences between those two samples, extracting the top N genes and use them for further evaluation.