[BioC] Re: Logit-t vs RMA

[Jenny Drnevich]

>> My enthusiasm regarding the paper is that it is the first one (to my
>> knowledge) that has interrogated probe_set differences at the probe
>> level  across groups (rather than pairwise),

I haven't had a chance to read this paper yet, but I am looking forward to
it...
However, have you seen: Chu, Weir, & Wolfinger.  A systematic statistical
linear modeling approach to oligonucleotide array experiments MATH BIOSCI
176 (1): 35-51 Sp. Iss. SI MAR 2002
They advocate using the probe-level data in a linear mixed model. 
Assuming that each probe is an independent measure (which I know is not
true because many of them overlap, but I'm ignoring this for now), using
probe-level data gives 14-20 "replicates" per chip. We've based our
analysis methods on this, and with two biological replicates per genetic
line, and three genetic lines per phenotypic group, we've been able to
detect as little as a 15% difference in gene expression at p=0.0001 (we
only expect 2 FP and get 60 genes with p=0.0001). Should have the
manuscript submitted in a couple weeks, and I'm working on using our
method on the benchmark data to see how it compares.  Results to be
discussed...

Cheers,
Jenny


especially since the
>> authors reported  what I suspected- that the probe level data does a
>> better job than the  expression values generated by a probe level
>> algorithm (I could still be  totally wrong, but it IS what I have been
>> suspecting). This may be more  important for discerning meaningful
>> differences than the transforms  themselves. If it is possible to
>> interrogate the data at this level, why  bother with probe level
>> algorithms which may lose information in the  process of cooking 11-32
>> intensity values into a single number?

-- 
Jenny Drnevich, Ph.D.
Department of Animal Biology
University of Illinois, Urbana-Champaign
515 Morrill Hall
505 S. Goodwin Ave.
Urbana, IL 61801
USA
ph: 217-244-6826
fax: 217-244-4565
drnevich at uiuc.edu