Costum Cdf Problem

Question

[cerikahp]

Hi!! Im new in working with costums CDF and im currently workin with MBNI's CDF. The thing is that with the normal CDF I obtain 1600 diferrentialy expresed gens, but when i I use the MBNI's CDF I get (with the same pipe line) 300, and that worries me. Im normalizing with rma, and the quality of the normalization and micro array seems okay.

Answer 1

[Houkto]

Hi, custom CDF will give you different results than affymetrix. However, since you are new to this there are different custom CDF (Transcript ID etc..) per chip. I have answered a question regarding custom CDF perhaps it will guide you on how to download custom CDF and point you to the journal which will explain why normal CDF will be different than Custom CDF. Custom Cdf For Affymetrix Human Genome U219 Array Plate?

[cerikahp]

thanks for answering, I was just a little worry how few diferentially expressed genes i have found. But playing with the normalization method let me see how much it can be affected.

[Houkto]

So the reason was the normalization method then ?

[cerikahp]

im not sure, i have to check the diferences in the anotation of the ones that didnt appear with costum CDF

[Houkto]

custom* not costum. when you downloaded custom CDF which type did you go for Entrezg, MIRBASEF, ENST, ENSG ? I am not sure what type of chip you have however try ENST or ENSG and see if you get a similar results ? also please accept an answer by clicking on the top arrow next to the answer section (if you think I have answered your question).

[cerikahp]

Entrez, but ENSG gives me similar results

[Houkto]

try ENST and see what happen ?

[cerikahp]

yep, 1350 transcripts mapping 400 genes

[Houkto]

then the difference will be 100 genes ?