Journal of Genetics

, Volume 89, Issue 1, pp 73–80

Reverse engineering large-scale genetic networks: synthetic versus real data

Research Article

DOI: 10.1007/s12041-010-0013-2

Cite this article as:
Zhang, L., Xiao, M., Wang, Y. et al. J Genet (2010) 89: 73. doi:10.1007/s12041-010-0013-2

Abstract

Development of microarray technology has resulted in an exponential rise in gene expression data. Linear computational methods are of great assistance in identifying molecular interactions, and elucidating the functional properties of gene networks. It overcomes the weaknesses of in vivo experiments including high cost, large noise, and unrepeatable process. In this paper, we propose an easily applied system, Stepwise Network Inference (SWNI), which integrates deterministic linear model with statistical analysis, and has been tested effectively on both simulated experiments and real gene expression data sets. The study illustrates that connections of gene networks can be significantly detected via SWNI with high confidence, when single gene perturbation experiments are performed complying with the algorithm requirements. In particular, our algorithm shows efficiency and outperforms the existing ones presented in this paper when dealing with large-scale sparse networks without any prior knowledge.

Keywords

gene regulatory networksingle gene perturbationlinear modelstepwisesimulated network

Copyright information

© Indian Academy of Sciences 2010

Authors and Affiliations

  1. 1.School of Computer Engineering and ScienceShanghai UniversityShanghaiPeople’s Republic of China
  2. 2.Institute of Systems BiologyShanghai UniversityShanghaiPeople’s Republic of China
  3. 3.Academy of Mathematics and Systems ScienceChinese Academy of Sciences (CAS)BeijingPeople’s Republic of China