The importance of platform annotation in interpreting microarray data.

摘要

Weigelt and colleagues1 recently compared three methods for classifying breast cancers into molecular subtypes. Aside from an obvious question of whether they should have centred each gene, an issue previously discussed in the context of classifying breast-cancer subtypes,2 there is concern that insufficient attention has been paid to the annotation of probes on the microarray platforms that were used. We review the PAM50 (pediction analysis of microarrays 50 gene set)3 SSP (single sample predictor), as applied to the Illumina BeadArray (Illumina Inc, San Diego, CA, USA) generated Natrajan data set.