Computational identification of proteins sub-network in Parkinson's disease study

Abstract

Parkinson's disease (PD) is a typical case of neurodegenerative disorder, which often impairs the sufferer's motor skills, speech, and other functions. Combination of proteinprotein interaction (PPI) network analysis and gene expression studies provides a better insight of Parkinson's disease. In our work a computational approach was developed to identify protein signal network in PD study. First, a network-constrain regularization analysis is employed to the linear regression model for gene expression data from transgenic mouse models in normal and with Parkinson's disease. Proteins sub-network was then detected based on an integer linear programming model by integrating microarray data and PPI database.