Processing Hg-U133A And Hg-U133B

Question

[drey]

Hello,

recently I found a data set that uses same experiment setup on these two platforms: HG-U133a and HG-U133b Since the overlap of probes is very small (168) I'm wondering if these platforms belong together. In other words: if it is a same expriment (lets say 50 samples), do I need to append probes in U133a (22283 probes) and U133b (22645 probes) to each other such that my data matrix will have a dimension of 44928x50 ? Or are these just completely "unrelated" platforms ?

Thank you

Answer 1

[Sean Davis]

I tend to treat them as completely unrelated platforms for normalization purposes. However, you'll want to merge them after normalization of you are going to be doing differential expression analysis since the FDR will depend on the total number of tests.

[drey]

Thank you! So for differential expression analysis, classification, and clustering I need to combine them.

[Sean Davis]

That is my suggestion, yes.

Answer 2

[Obi Griffith]

They are certainly not unrelated. They were originally designed and marketed as the Affymetrix U133 set. "The HG-U133A Array includes representation of the RefSeq database sequences and probe sets related to sequences previously represented on the Human Genome U95Av2 Array. The HG-U133B Array contains primarily probe sets representing EST clusters." So, they were meant to be complementary and non-overlapping. However, due to the fact that the most well known genes were included on the U133A array, many studies have opted to use only the U133A set.