PMINR: Pointwise Mutual Information-Based Network Regression – With Application to Studies of Lung Cancer and Alzheimer’s Disease

Abstract

Complex diseases are believed to be the consequence of intracellular network(s) involving a range of factors. An improved understanding of a disease-predisposing biological network could lead to better identification of genes and pathways that confer disease risk and therefore inform drug development. The group difference in biological networks, as is often characterized by graphs of nodes and edges, is attributable to effects of these nodes and edges. Here we introduced pointwise mutual information (PMI) as a measure of the connection between a pair of nodes with either a linear relationship or nonlinear dependence. We then proposed a PMI-based network regression (PMINR) model to differentiate patterns of network changes (in node or edge) linking a disease outcome. Through simulation studies with various sample sizes and inter-node correlation structures, we showed that PMINR can accurately identify these changes with higher power than current methods and be robust to the network topology. Finally, we illustrated, with publicly available data on lung cancer and gene methylation data on aging and Alzheimer's disease, an evaluation of the practical performance of PMINR. We concluded that PMI is able to capture the generic inter-node correlation pattern in biological networks, and PMINR is a powerful and efficient approach for biological network analysis.

Network regression analysis in transcriptome-wide association studies

“…It should be noted that the link may be nonlinear. We have previously proposed PMINR [ 22 ] for efficient network regression analysis, where pointwise mutual information (PMI) is used to measure the strength of the connection between a pair of nodes. PMINR has shown better performance in capturing the general relationship among different nodes in a biological network than other methods including PMNR [ 22 ], DGCA [ 23 ] and RANK [ 24 ].…”

“…We have previously proposed PMINR [ 22 ] for efficient network regression analysis, where pointwise mutual information (PMI) is used to measure the strength of the connection between a pair of nodes. PMINR has shown better performance in capturing the general relationship among different nodes in a biological network than other methods including PMNR [ 22 ], DGCA [ 23 ] and RANK [ 24 ]. Specifically, PMNR uses the common linear correlation to represent the between-node connection strength for network regression [ 22 ], DGCA is differential gene correlation analysis (i.e., edge effect) to assess the difference in gene regulatory relationships under multiple conditions [ 23 ], while RANK can detect the whole pathway due to either correlation or mean changes [ 24 ].…”

“…PMINR has shown better performance in capturing the general relationship among different nodes in a biological network than other methods including PMNR [ 22 ], DGCA [ 23 ] and RANK [ 24 ]. Specifically, PMNR uses the common linear correlation to represent the between-node connection strength for network regression [ 22 ], DGCA is differential gene correlation analysis (i.e., edge effect) to assess the difference in gene regulatory relationships under multiple conditions [ 23 ], while RANK can detect the whole pathway due to either correlation or mean changes [ 24 ]. However, the modeling framework of PMINR requires all the gene network nodes to be observed, thus cannot be directly implemented for network regression in TWAS analysis, where the gene expression are commonly unobserved in the GWAS.…”

Mechanism of Zhinao Capsule in Treating Alzheimer’s Disease Based on Network Pharmacology Analysis and Molecular Docking Validation

“…Research suggests that miR-132 may improve the cognition of rats with AD by inhibiting the MAPK signaling pathway [ 55 ]. Furthermore, MAPK1 is related to the hyperphosphorylation of Tau protein, which is considered an important factor leading to the neuropathological changes in AD [ 56 ]. Another key target is the cytokine TNF- α , which is involved in inflammation throughout the body [ 57 ].…”

Differential co-expression network analysis with DCoNA reveals isomiR targeting aberrations in prostate cancer

“…The studies [11, 13] categorized pairs into all possible paired correlation scenarios, that among other things, allowed us to identify pairs that experience no correlation in one condition but become correlated in another. As an example of non-correlation methods, we mention PMINR [14], where authors used mutual information and build a regression model to detect pairwise interactions related to a disease.…”