Transformations…
what for?
which one?

Slide 2

u ratios and fold changes

u ratios and fold changes

u ratios and fold changes

u In the following:

u Sources of variation

Slide 8

u The two-component model

u variance stabilizing transformations

u variance stabilizing transformations

u variance stabilizing transformations

u the “glog” transformation

parameter estimation

Least trimmed sum of squares regression

evaluation: effects of different data transformations

u Normality: QQ-plot

evaluation: sensitivity / specificity in
detecting differential abundance

evaluation: comparison of methods

evaluation: a benchmark for Affymetrix genechip expression measures

u affycomp results (28 Sep 2003)

u ROC curves

u Stratification

u glog versus "sliding z-score"

u Availability

u What to do with the gene lists:
the functional genomics pipeline @ DKFZ

u HT functional assays (S. Wiemann, D. Arlt)

Slide 28

Slide 29

u Cellular assays: challenges for statisticians

u Summary

u Acknowledgements

u Models are never correct, but some are useful

Slide 34

u ratio compression