Hypergraph models of biological networks to identify genes critical to pathogenic viral response.

BACKGROUND: Representing biological networks as graphs is a powerful approach to reveal underlying patterns, signatures, and critical components from high-throughput biomolecular data. However, graphs do not natively capture the multi-way relationships present among genes and proteins in biological systems. Hypergraphs are generalizations of graphs that naturally model multi-way relationships and have shown promise in modeling systems such as protein complexes and metabolic reactions. In this paper we seek to understand how hypergraphs can more faithfully identify, and potentially predict, important genes based on complex relationships inferred from genomic expression data sets. RESULTS: We compiled a novel data set of transcriptional host response to pathogenic viral infections and formulated relationships between genes as a hypergraph where hyperedges represent significantly perturbed genes, and vertices represent individual biological samples with specific experimental conditions. We find that hypergraph betweenness centrality is a superior method for identification of genes important to viral response when compared with graph centrality. CONCLUSIONS: Our results demonstrate the utility of using hypergraphs to represent complex biological systems and highlight central important responses in common to a variety of highly pathogenic viruses.

Mathematical methods for visualization and anomaly detection in telemetry datasets.

Recent developments in both biological data acquisition and analysis provide new opportunities for data-driven modelling of the health state of an organism. In this paper, we explore the evolution of temperature patterns generated by telemetry data collected from healthy and infected mice. We investigate several techniques to visualize and identify anomalies in temperature time series as temperature relates to the onset of infectious disease. Visualization tools such as Laplacian Eigenmaps and Multidimensional Scaling allow one to gain an understanding of a dataset as a whole. Anomaly detection tools for nonlinear time series modelling, such as Radial Basis Functions and Multivariate State Estimation Technique, allow one to build models representing a healthy state in individuals. We illustrate these methods on an experimental dataset of 306 Collaborative Cross mice challenged with Salmonella typhimurium and show how interruption in circadian patterns and severity of infection can be revealed directly from these time series within 3 days of the infection event.