Master Regulators Used as Breast Cancer Metastasis Classifier

Abstract

Computational identification of prognostic biomarkers capable of withstanding follow-up validation efforts is still an open challenge in cancer research. For instance, several gene expression profiles analysis methods have been developed to identify gene signatures that can classify cancer sub-phenotypes associated with poor prognosis. However, signatures originating from independent studies show only minimal overlap and perform poorly when classifying datasets other than the ones they were generated from. In this paper, we propose a computational systems biology approach that can infer robust prognostic markers by identifying upstream Master Regulators, causally related to the presentation of the phenotype of interest. Such a strategy effectively extends and complements other existing methods and may help further elucidate the molecular mechanisms of the observed pathophysiological phenotype. Results show that inferred regulators substantially outperform canonical gene signatures both on the original dataset and across distinct datasets.

The Reconstruction of Transcriptional Networks Reveals Critical Genes with Implications for Clinical Outcome of Multiple Myeloma

“…We found that genes with relatively low numbers of connections impaired network stability in all the data sets analyzed. This finding is consistent with previous evidence suggesting that reverse engineering approaches are able to discover master regulator genes with limited numbers of interactions largely affecting the transcriptome (11,12,26). Table 2 and Supplementary Table S1 list the top 100 genes with the highest GCS derived from our analysis (the whole criticality scores for each data set are provided as Supplementary Table S2).…”

“…Moreover, network theory-derived procedures have been developed that describe the internal stability of complex networks as resilience to perturbation, which can be used to identify genes that have a critical role in the stability and functioning of the network (10). This approach allows to define master regulator genes that might explain the biology of disease subsets (11)(12)(13). The selection of such genes, based on their role in a critical regulatory network, rather than solely on their level of expression, may improve the reliability and robustness of expression analysis to predict outcome.…”

Regulation of extra-embryonic endoderm stem cell differentiation by Nodal and Cripto signaling

“…Redundant probes were eliminated by keeping the one showing the Development 138 (18) DEVELOPMENT highest dynamic range (coefficient of variation) among samples. False discovery rate (FDR) was estimated as described (Benjamini and Hochberg, 1995 (Lim et al, 2009)]. Briefly, we divided the query gene set into two subsets: a positive subset containing the upregulated part of the query signature, and a negative subset encompassing the downregulated part of the query signature.…”

“…Indeed, GSEA is the most effective approach to test whether two processes produce similar effects in terms of differentially expressed genes (Subramanian et al, 2005). Using an enhanced version of this algorithm, termed two-tails GSEA, which takes into account the 'direction' of the gene expression change (Lim et al, 2009), we found a strong enrichment of the Nodal signature on the Cripto gene expression profile (Fig. 3C) and vice versa (Fig.…”

“…Enrichment of Gene Ontology Biological Process (GO-BP) gene sets on Nodal and Cripto gene expression signatures were assessed by gene set enrichment analysis [GSEA (Subramanian et al, 2005)]. Statistical significance for the overlap between signatures was computed by Fisher's exact test and by a modified version of GSEA [two-tails GSEA (Lim et al, 2009)]. Briefly, we divided the query gene set into two subsets: a positive subset containing the upregulated part of the query signature, and a negative subset encompassing the downregulated part of the query signature.…”