Structural and dynamic analysis of biological networks

Date
2015
Authors
Mohyedin Bonab, Elmira
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract

Network analytic provides a systematic approach in revealing the complexity of biological processes. Consequently, the statistical and structural properties of these biological networks are quantitatively comparable with other well-studied networks such as social networks. We constructed three application-dependent biological networks: a gene regulatory network, signaling network and cold-stimulated white adipose gene network. The gene regulatory network is constructed using correlations between temporal gene expression profiles along with molecular binding information. The signaling network is built using human signaling transduction pathways. In the latter network, differentially expressed cold-induced genes are connected as they participate in similar biological processes and have significantly correlated expression profiles. Information spreading patterns or activation patterns in a biological network reveals network modularity features. There is a dependency between where the flow is originated and where is traveling in molecular activation patterns. Therefore, we compared the ability of a random walker, as the simplest information spreading modeler, with second-order Markov dynamics in biological networks. As compared to first-order Markov dynamics (random walker), second-order Markov model can better identify connections between chains of biological processes in signaling and white adipose tissue networks, such that detected modules by second-order Markov dynamics have distinctive, specific and densely interrelated biological processes.

Description
This item is available only to currently enrolled UTSA students, faculty or staff.
Keywords
Markov, Andrey, Gene network
Citation
Department
Electrical and Computer Engineering