Non-linear dimensionality reduction of signaling networks

Sergii Ivakhno,J Douglas Armstrong

doi:10.1186/1752-0509-1-27

Abstract

BackgroundSystems wide modeling and analysis of signaling networks is essential for understanding complex cellular behaviors, such as the biphasic responses to different combinations of cytokines and growth factors. For example, tumor necrosis factor (TNF) can act as a proapoptotic or prosurvival factor depending on its concentration, the current state of signaling network and the presence of other cytokines. To understand combinatorial regulation in such systems, new computational approaches are required that can take into account non-linear interactions in signaling networks and provide tools for clustering, visualization and predictive modeling.ResultsHere we extended and applied an unsupervised non-linear dimensionality reduction approach, Isomap, to find clusters of similar treatment conditions in two cell signaling networks: (I) apoptosis signaling network in human epithelial cancer cells treated with different combinations of TNF, epidermal growth factor (EGF) and insulin and (II) combination of signal transduction pathways stimulated by 21 different ligands based on AfCS double ligand screen data. For the analysis of the apoptosis signaling network we used the Cytokine compendium dataset where activity and concentration of 19 intracellular signaling molecules were measured to characterise apoptotic response to TNF, EGF and insulin. By projecting the original 19-dimensional space of intracellular signals into a low-dimensional space, Isomap was able to reconstruct clusters corresponding to different cytokine treatments that were identified with graph-based clustering. In comparison, Principal Component Analysis (PCA) and Partial Least Squares – Discriminant analysis (PLS-DA) were unable to find biologically meaningful clusters. We also showed that by using Isomap components for supervised classification with k-nearest neighbor (k-NN) and quadratic discriminant analysis (QDA), apoptosis intensity can be predicted for different combinations of TNF, EGF and insulin. Prediction accuracy was highest when early activation time points in the apoptosis signaling network were used to predict apoptosis rates at later time points. Extended Isomap also outperformed PCA on the AfCS double ligand screen data. Isomap identified more functionally coherent clusters than PCA and captured more information in the first two-components. The Isomap projection performs slightly worse when more signaling networks are analyzed; suggesting that the mapping function between cues and responses becomes increasingly non-linear when large signaling pathways are considered.ConclusionWe developed and applied extended Isomap approach for the analysis of cell signaling networks. Potential biological applications of this method include characterization, visualization and clustering of different treatment conditions (i.e. low and high doses of TNF) in terms of changes in intracellular signaling they induce.

Highlights

Systems wide modeling and analysis of signaling networks is essential for understanding complex cellular behaviors, such as the biphasic responses to different combinations of cytokines and growth factors
To analyze the apoptosis network we considered a previously published protein signaling dataset known as the Cytokine compendium [6], for which quantitative western blotting, high-throughput protein kinase assays and protein microarrays were used to investigate the combinatorial effect of tumor necrosis factor (TNF), epidermal growth factor (EGF) and insulin on apoptosis of human adenocarcinoma cells
HT29 epithelial cancer cells were treated with 10 combinations of saturating or subsaturating concentrations of TNF, EGF and insulin (0, 0.2, 5, 100 ng/ml TNF and 0, 1, 100 ng/ml EGF or 0, 1, 5, 500 ng/ml insulin respectively), which collectively represent all the cues used in the study. 19 molecular signals were chosen to characterize changes in signaling network activity in response to each cue condition [see Additional file 1]

Summary

Introduction

Systems wide modeling and analysis of signaling networks is essential for understanding complex cellular behaviors, such as the biphasic responses to different combinations of cytokines and growth factors. Traditional biochemical and molecular biology approaches focus on functional contributions of individual molecules and often overlook that activation of multiple signaling molecules determines cellular response to external stimuli Such information will be crucial for constructing predictive and descriptive models of cell decision processes and identifying how pathological conditions such as uncontrollable cellular proliferation arise from abnormalities in signaling networks. In contrast to transcriptional regulatory networks for which high-throughput technologies such as DNA microarrays, genome-wide knockouts/RNAi and chromatin-immunoprecipitation are available for profiling network activities, the analysis of signaling networks at the biochemical/ molecular level has been hindered by the absence of sensitive high-throughput approaches. Several such methodologies have been described that use either multiparameter flow cytometry with causal (Bayesian) networks [1,2] or combinations of protein signaling assays with Bayesian networks [3], Principal Components Analysis (PCA) [4], and Partial Least-Squares Regression (PLSR) [5]

Objectives

Methods

Results

Discussion

Conclusion