A Combined Analysis of Genome-Wide Expression Profiling of Bipolar Disorder in Human Prefrontal Cortex.

Jinglu Wang,Susu Qu,Weixiao Wang,Liyuan Guo,Kunlin Zhang,Suhua Chang,Jing Wang

DOI: https://doi.org/10.1016/j.jpsychires.2016.07.016

IF: 5.25

2016-01-01

Journal of Psychiatric Research

Abstract:Numbers of gene expression profiling studies of bipolar disorder have been published. Besides different array chips and tissues, variety of the data processes in different cohorts aggravated the inconsistency of results of these genome-wide gene expression profiling studies. By searching the gene expression databases, we obtained six data sets for prefrontal cortex (PFC) of bipolar disorder with raw data and combinable platforms. We used standardized pre-processing and quality control procedures to analyze each data set separately and then combined them into a large gene expression matrix with 101 bipolar disorder subjects and 106 controls. A standard linear mixed-effects model was used to calculate the differentially expressed genes (DEGs). Multiple levels of sensitivity analyses and cross validation with genetic data were conducted. Functional and network analyses were carried out on basis of the DEGs. In the result, we identified 198 unique differentially expressed genes in the PFC of bipolar disorder and control. Among them, 115 DEGs were robust to at least three leave-one-out tests or different pre-processing methods; 51 DEGs were validated with genetic association signals. Pathway enrichment analysis showed these DEGs were related with regulation of neurological system, cell death and apoptosis, and several basic binding processes. Protein-protein interaction network further identified one key hub gene. We have contributed the most comprehensive integrated analysis of bipolar disorder expression profiling studies in PFC to date. The DEGs, especially those with multiple validations, may denote a common signature of bipolar disorder and contribute to the pathogenesis of disease.