Reconstruction of Gene Regulatory Networks Using Principal Component Analysis
Document Type
Book Chapter
Publication Date
3-22-2016
School
Computing Sciences and Computer Engineering
Abstract
Reconstructing complex networks of genetic interactions and identifying unknown relations among genes are very important for understanding the underlying mechanism of biological processes. Many inference methods have been proposed to reconstruct unknown gene regulatory networks (GRNs) using microarray data sets. A state space model (SSM) is a method that can be used to infer GRNs from a time series data set. To overcome these difficulties in SSM, we introduce a simplified linear model (SLM) which uses principal component analysis (PCA) to reduce the dimension of the data set and the noise. The performance of SSM and SLM was evaluated in terms of accuracy, efficiency, and sensitivity to the time series data sets. Results show that SLM can be applied to a smaller data set and is more stable than SSM with different lengths of hidden variables. Additionally, the proposed SLM is simpler and more computationally efficient than SSM without sacrificing accuracy.
Publication Title
Emerging Trends in Applications and Infrastructures for Computational Biology, Bioinformatics, and Systems Biology: Systems and Applications
First Page
171
Last Page
179
Recommended Citation
Wu, X.,
Yang, B.,
Maxwell, A.,
Koh, W.,
Gong, P.,
Zhang, C.
(2016). Reconstruction of Gene Regulatory Networks Using Principal Component Analysis. Emerging Trends in Applications and Infrastructures for Computational Biology, Bioinformatics, and Systems Biology: Systems and Applications, 171-179.
Available at: https://aquila.usm.edu/fac_pubs/19544
COinS