New connections: Making discoveries in complex data sets

See allHide authors and affiliations

Science Signaling  07 Jun 2016:
Vol. 9, Issue 431, pp. ec132
DOI: 10.1126/scisignal.aag2863

Cells are never exposed to only one signal at a time. They are bathed in a complex, dynamically changing milieu of growth factors, nutrients, cytokines, and hormones, which creates enormous complexity for studying cellular regulation. The complexity of this situation means that computational approaches are needed to find meaning in large, complex, and diverse data sets. Chitforoushzadeh et al. applied a statistical modeling approach called “tensor partial least squares regression,” which maintains data structures as multidimensional elements called tensors. Transcriptomic and proteomic (signaling protein abundance, phosphorylation status, and activity) data were assessed with tensor modeling to explore how signaling by the growth factors insulin and EGF integrated with that of the proinflammatory cytokine tumor necrosis factor–α (TNF-α). This analysis revealed a specific phosphorylation event on the long form of the transcription factor GATA that enabled insulin to inhibit the expression of genes targeted by TNF-α. Gierut et al. also investigated TNF-α signaling, this time in the context of the induction of cell death in the intestinal epithelium of mice pretreated with various kinase inhibitors and then administered TNF-α. The goal was to gain insight into how various phosphorylation-dependent signaling events contributed to cell survival or apoptosis. Unexpectedly, inhibitors that targeted the same kinase with equal efficiencies had different effects on TNF-α–induced apoptosis. The mathematical approach used to reveal an unexpected molecular survival signal was discriminant partial least squares regression (D-PLSR) analysis with Monte Carlo subsampling followed by cluster analysis of a self-organizing map (SOM). This analysis revealed that the kinase Akt, which usually promotes cell survival, increased TNF-α–induced apoptosis when the network had been altered by the presence of a mitogen-activated protein kinase kinase (MEK) inhibitor that enhanced the death response. SOM analysis was the approach used by Tsigelny et al. to uncover genes that orchestrate the transitions between stages of kidney development. By collecting more than 30,000 genes into 650 groups called metagenes and applying SOM and entropy analysis, the authors correlated metagene-defined stages with morphometric parameters and specific gene networks, which resulted in the identification of known and previously unknown genes involved in the process of kidney development. These three studies illustrate the power of computational approaches to generate testable predictions and provide insight into the complexity of cellular regulatory networks.

Z. Chitforoushzadeh, Z. Ye, Z. Sheng, S. LaRue, R. C. Fry, D. A. Lauffenburger, K. A. Janes, TNF-insulin crosstalk at the transcription factor GATA6 is revealed by a model that links signaling and transcriptomic data tensors. Sci. Signal. 9, ra59 (2016). [Abstract]

J. J. Gierut, L. B. Wood, K. S. Lau, Y.-J. Lin, C. Genetti, A. A. Samatar, D. A. Lauffenburger, K. M. Haigis, Network-level effects of kinase inhibitors modulate TNF-α–induced apoptosis in the intestinal epithelium. Sci. Signal. 8, ra129 (2015). [Abstract]

I. F. Tsigelny, V. L. Kouznetsova, D. E. Sweeney, W. Wu, K. T. Bush, S. K. Nigam, Analysis of metagene portraits reveals distinct transitions during kidney organogenesis. Sci. Signal. 1, ra16 (2008). [Abstract]