Removing cell cycle genes from signatures does not eliminate the confounding effect of cell proliferationproliferation on outcome association. This indicates that proliferation effects extend beyond classical cell cycle gene expression and cannot be adequately controlled through selective gene exclusion. [@venet_most_2011]

Definitions

Synthesis

Despite efforts to remove cell cycle genes from prognostic signatures, proliferation confounding persists in gene expression-outcome associations, suggesting that the influence of proliferation extends beyond directly annotated cell cycle genes. The mechanistic basis for this persistence likely involves the fact that proliferation signals permeate many biological processes and gene expression patterns, creating a confounding effect that cannot be eliminated simply by excluding genes with explicit cell cycle annotations. This raises an unresolved question about whether standard approaches to controlling for proliferation are fundamentally inadequate, and whether the statistical associations observed between gene signatures and clinical outcomes reflect genuine biological mechanisms or merely correlate with the underlying proliferative state of tumors.

Bibliography