In 2009, we reported a clinically applicable gene expression-based predictor that robustly identifies these main intrinsic subtypes by quantitative measurement of 50 genes (i.e., PAM50) [ 1].
Over the last decade, global gene expression profiling has given us insights into the biological complexity of breast tumors, and clinically applicable gene expression-based assays are being developed for the prediction of prognosis and/or treatment benefit [ 6– 14].
The method is also directly applicable to gene expression and proteomics.
While not demonstrated here, the method is immediately applicable to gene expression data generated with sequencing technologies (RNA-seq) once gene-level measurements have been derived.
Another approach to finding functionally relevant groups of genes is network derivation, which has been popular in the analysis of gene-gene and protein-protein interactions [ 6- 10], and is also applicable to gene expression analysis [ 11, 12].
Briefly, PAM is an efficient tool to identify molecular patterns that is applicable to gene expression data, and uses nearest shrunken centroids to identify subsets of genes that best characterize sample classes.
NMF is an efficient method to identify molecular patterns that is readily applicable to gene expression data, and therefore can be used as a powerful means for class discovery.
An estimate of the overall proportion of differentially accumulating transcripts (1-π0) suggesthatthas as many as 18% of the genes may be differentially expressed between B and M. The analytical model presented here is generally applicable to gene expression data and demonstrates the use of statistical elimination of confounding effects such as stress in the context of microarray analysis.
Finally, to test whether our quantile-based scaling is applicable to gene expression intensities derived from other tissue, we processed RNA from two pluripotent human embryonic stem cell (hESC) lines (H1 and H9; Additional file 3) using both microarray and RNA-Seq, which were used to assess, but not to improve, microarray quality.
We consider the GM call to be a true, scale free, normalization that is entirely platform-independent, applicable to any gene expression.
