Recent advances in high-throughput technology and the emergence of large-scale genomic datasets have enabled detection of genomic features that affect clinical outcomes. Although many previous computational studies have analysed the effect of each single gene or the additive effects of multiple genes on the clinical outcome, less attention has been devoted to the identification of gene-gene interactions of general type that are associated with the clinical outcome. Moreover, the integration of information from multiple molecular profiles adds another challenge to this problem. Recently, network-based approaches have gained huge popularity. However, previous network construction methods have been more concerned with the relationship between features only, rather than the effect of feature interactions on clinical outcome. We propose a mutual information-based integrative network analysis framework (MINA) that identifies gene pairs associated with clinical outcome and systematically analyses the resulting networks over multiple genomic profiles. We implement an efficient non-parametric testing scheme that ensures the significance of detected gene interactions. We develop a tool named MINA that automates the proposed analysis scheme of identifying outcome-associated gene interactions and generating various networks from those interacting pairs for downstream analysis. We demonstrate the proposed framework using real data from ovarian cancer patients in The Cancer Genome Atlas (TCGA). Statistically significant gene pairs associated with survival were identified from multiple genomic profiles, which include many individual genes that have weak or no effect on survival. Moreover, we also show that integrated networks, constructed by merging networks from multiple genomic profiles, demonstrate better topological properties and biological significance than individual networks. We have developed a simple but powerful analysis tool that is able to detect gene-gene interactions associated with clinical outcome on multiple genomic profiles. By being network-based, our approach provides a better insight into the underlying gene-gene interaction mechanisms that affect the clinical outcome of cancer patients.

Aspirin Exacerbated Respiratory Disease (AERD) is a chronic medical condition that encompasses asthma, nasal polyposis, and hypersensitivity to aspirin and other non-steroidal anti-inflammatory drugs. Several previous studies have shown that part of the genetic effects of the disease may be induced by the interaction of multiple genetic variants. However, heavy computational cost as well as the complexity of the underlying biological mechanism has prevented a thorough investigation of epistatic interactions and thus most previous studies have typically considered only a small number of genetic variants at a time. In this study, we propose a gene network based analysis framework to identify genetic risk factors from a genome-wide association study dataset. We first derive multiple single nucleotide polymorphisms (SNP)-based epistasis networks that consider marginal and epistatic effects by using different information theoretic measures. Each SNP epistasis network is converted into a gene-gene interaction network, and the resulting gene networks are combined as one for downstream analysis. The integrated network is validated on existing knowledgebase of DisGeNET for known gene-disease associations and GeneMANIA for biological function prediction. We demonstrated our proposed method on a Korean GWAS dataset, which has genotype information of 440,094 SNPs for 188 cases and 247 controls. The topological properties of the generated networks are examined for scale-freeness, and we further performed various statistical analyses in the Allergy and Asthma Portal (AAP) using the selected genes from our integrated network. Our result reveals that there are several gene modules in the network that are of biological significance and have evidence for controlling susceptibility and being related to the treatment of AERD.

Network-based approaches have recently gained considerable popularity in high- dimensional regression settings. For example, the Cox regression model is widely used in expression analysis to predict the survival of patients. However, as the number of genes becomes substantially larger than the number of samples, the traditional Cox or L2-regularized Cox models are still prone to noise and produce unreliable estimations of regression coefficients. A recent approach called the network-based Cox (Net-Cox) model attempts to resolve this issue by incorporating prior gene network information into the Cox regression. The Net-Cox model has shown to outperform the models that do not use this network information.