1 CRIBI
Biotecnology Centre, University of Padova, Italy
2Department of Statistical Sciences, University of Padova,
via C. Battisti 241, 35121 Padova, Italy
Motivation. Microarray normalization is a fundamental step in removing systematic
bias and noise variability caused by technical and experimental
artefacts. Several approaches, suitable for large-scale genome arrays,
have been proposed and shown to be effective in the reduction of
systematic errors. Most of these methodologies are based on specific
assumptions that are reasonable for whole-genome arrays but possibly
unsuitable for small microRNA platforms. The aim of this work is the
investigation, through simulated and real datasets, of the influence
that normalizations for two-colour microRNA arrays have on the
identification of differentially expressed genes.
Results. We show that normalizations usually applied to
large-scale arrays, in several cases, completely modify the actual
structure of microRNA data, leading to large portions of false positives
and false negatives. Here, we propose a novel normalization (loessM),
based on loess algorithm, that scales data on the median expression
values. LoessM is able to outperform other techniques in most
experimental scenarios. Moreover, when usual assumptions on
differentially expression distribution are missed, channel-effect has a
strikingly negative influence on small arrays, bias that cannot be
removed by normalizations but rather by an appropriate experimental
design. We find that the combination of loessM with eCADS, an
experimental design based on biological replicates dye-swap recently
proposed for channel-effect reduction, gives the best results in term of
specificity/sensitivity either on simulated and on real data.