[BioC] AffyRNAdeg slopes

James W. MacDonald jmacdon at med.umich.edu
Mon Dec 6 20:46:56 CET 2004

Vawter, Marquis wrote:
> Hi Jim, just a follow-up to your post, when you referred to outliers on the
> density plot in discussing AffyRNAdeg slopes, is this a function within the
> AffyRNAdeg program?  Best wishes, Mark

No, it's just an eyeballometric measurement on my part. However, it 
appears that you might be confusing what I said. My point was that the 
results from the density plot (different looking curves) are usually 
much more indicative of how things are going to work out than the 
AffyRNAdeg plots (non-parallel lines).

When I analyze Affy data, I always do the density and AffyRNAdeg plots 
as QC. If I have a chip or chips that look different (on the density 
plot), I will use the residual plots in the affyPLM package to see how 
well the medianpolish algorithm is fitting the data, and the 'outliers' 
usually have huge residuals.

Another thing I like to do to QC the data is to do a principal 
components analysis of the expression values (e.g., after fitting RMA or 
GCRMA or whatever) and plot the first two principal components. If I 
don't get reasonable grouping of my samples on this plot it usually 
indicates that the replicates are not very similar, which almost never 
bodes well for the downstream analyses.


> "Anyway, in my experience, large differences in slope are less critical than
> big differences between the density plots. I find that outliers on the
> density plot nearly always require a redo of the chip, whereas chips with
> very different slopes in the RNA degradation plot may be usable."
> Marquis P. Vawter, Ph.D.
> Director, Functional Genomics Laboratory
> Department of Psychiatry & HB
> Medical Science - D433, Zot Code 1675
> University of California, College of Medicine 
> Irvine CA 92697-1675
> Email mvawter at uci.edu <mailto:mvawter at uci.edu>
> Phone 949-824-9014
> Fax 949-824-7012
> Lab: Gillespie Neuroscience Bldg., Room 3226
> Zot code 4260B.
> Lab: 949-824-9170
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor

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

More information about the Bioconductor mailing list