[BioC] Top 50 Candidate genes: "wrong" pValues and B results!

[James MacDonald]

Adrian,

It is quite possible to get this sort of result with microarray data.
First, if you only have a few replicates, the power of a t-test is
pretty low so it is difficult to get any large t-statistics (although I
have no idea how many replicates you have). Second, the default
multiplicity adjustment for topTable is the Holm procedure, which is
fairly conservative (you don't say what you used...). Third, you only
have six genes in this list with a fold change greater than two-fold
anyway, which indicates that there may not be any real big differences
between the two samples. Note that there may be more genes with FC > 2
farther down the list, but since these will have smallish t-statistics,
the fold change is likely due to an outlier or two.

You may simply need to do some more microarrays in order to get enough
power to detect differentially expressed genes. Alternatively, you might
just try validating some of the top genes in this list and see if you
see any differential expression using RTPCR or Northern blots.

Best,

Jim



James W. MacDonald
Affymetrix and cDNA Microarray Core
University of Michigan Cancer Center
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109
734-647-5623

>>> Adrian Peres <adrian.peres at cropdesign.com> 07/26/04 03:47AM >>>
Hi all
I am using Limma GUI to analyze Agilent data and I have some troubles.
Hereafter I placed a sample of the statistic analysis  that looks very

strange to me!
The pValues are all "1" and I don't understand why? B data are not 
better neither...
Can my data be so bad?
Greetings,
Adrian
M 	A 	t 	Pvalue 	B
2.099 	9.493 	27.16 	1 	-3.417
-1.389 	11.37 	-15.45 	1 	-3.437
-1.079 	8.683 	-15.28 	1 	-3.438
-1.073 	8.952 	-14.28 	1 	-3.442
0.944 	6.831 	12.66 	1 	-3.451
-0.8108 	8.475 	-11.51 	1 	-3.46
-0.7985 	8.784 	-11.39 	1 	-3.461
-0.803 	12.67 	-11.38 	1 	-3.461
-0.9375 	12.22 	-11.13 	1 	-3.463
-0.8701 	10.24 	-10.84 	1 	-3.466
-1.383 	12.55 	-10.63 	1 	-3.468
-0.846 	9.56 	-10.22 	1 	-3.473
-0.7271 	11.98 	-10.1 	1 	-3.474
0.977 	8.603 	9.975 	1 	-3.476
-0.7734 	9.416 	-9.716 	1 	-3.48
-0.975 	9.89 	-9.52 	1 	-3.482
-0.6683 	7.127 	-9.514 	1 	-3.482
-0.8116 	6.528 	-9.493 	1 	-3.483
-0.673 	8.966 	-9.434 	1 	-3.484
-0.8475 	11.26 	-9.125 	1 	-3.489
-0.8426 	10.43 	-9.038 	1 	-3.49
-0.7633 	8.999 	-8.977 	1 	-3.491
-0.8202 	10.14 	-8.804 	1 	-3.494
-0.6119 	9.076 	-8.671 	1 	-3.497
-0.7045 	9.458 	-8.611 	1 	-3.498
-0.761 	10.07 	-8.468 	1 	-3.501
-0.9212 	10.8 	-8.437 	1 	-3.502
-1.003 	8.465 	-8.392 	1 	-3.502


confidentiality notice:
The information contained in this e-mail is confidential\ an...{{dropped}}