[BioC] affy technical vs biological replicates analysis

jane janes reymonnn at gmail.com
Tue May 29 10:43:30 CEST 2007

Dear list,

I have 24 chicken affymetrix arrays.

Half of them are technical replicates (2 different conditions, 3 time
points per condition (2X3=6 RNA samples) each RNA sample was analysed
twice (6X2=12 microarrays).

The other 12 samples come from 6 different conditions, two biological
replicates (different RNA samples) per condition.

Up to now I have been analyzing the results only from the first 12
samples using webbioc (RMA processing, t - test (equal variance) for
pairwise comparison and F - statistics for different time points)
considering each different microarray as a different experiment.
(the results for top differentially expressed genes were quite good
and in agreement with already published data)

Now I would like to look for differentially expressed genes among the
first 12 samples (2 different conditions, 3 time points per condition)
as well to compare some of them with samples coming from the
biological replicates.

Here come and my questions.
1. How shell I treat my technical duplicates in term to compare them
with biological duplicates.
Any example code or link to such will be greatly appreciated.
2. Having only duplicates (technical or biological) am I allowed to
use t-statistics (equal variance) with webbioc/multtest or I can use
only limma's linear models and eBayes.

Thank you in advance for your answers,

Best regards,

Vassil Vassilev
CDB, Japan

More information about the Bioconductor mailing list