|
Abstract:
|
DNA microarray technology measures hybridization of nucleic acids to complementary molecules attached to a solid surface, referred to as probes. This high-content genomic tool typically generates many measurements of which only a relatively small subset is informative for the interpretation of the experiment. Here, we show how a probe set can be called informative or non-informative using probe level information. The underlying idea is that probes of an informative probe set detect the mRNA transcript of this probe set rather consistently if its mRNA concentration varies across arrays. This gene selection approach outperforms currently available methods, and is entirely objective. Consequently, by excluding many false positives, it offers a key solution to the main problems in the analysis of high-dimensional microarray data, being multiple testing and overfitting.
|
