Deciphering deterioration mechanisms of complex diseases based on the construction of dynamic networks and systems analysis

Abstract

The early diagnosis and investigation of the pathogenic mechanisms of complex diseases are the most challenging problems in the fields of biology and medicine. Network-based systems biology is an important technique for the study of complex diseases. The present study constructed dynamic protein-protein interaction (PPI) networks to identify dynamical network biomarkers (DNBs) and analyze the underlying mechanisms of complex diseases from a systems level. We developed a model-based framework for the construction of a series of time-sequenced networks by integrating high-throughput gene expression data into PPI data. By combining the dynamic networks and molecular modules, we identified significant DNBs for four complex diseases, including influenza caused by either H3N2 or H1N1, acute lung injury and type 2 diabetes mellitus, which can serve as warning signals for disease deterioration. Function and pathway analyses revealed that the identified DNBs were significantly enriched during key events in early disease development. Correlation and information flow analyses revealed that DNBs effectively discriminated between different disease processes and that dysfunctional regulation and disproportional information flow may contribute to the increased disease severity. This study provides a general paradigm for revealing the deterioration mechanisms of complex diseases and offers new insights into their early diagnoses.

From feedback loop transitions to biomarkers in the psycho-immune-neuroendocrine network: Detecting the critical transition from health to major depression

“…While critical transitions generally occur at a 'whole-of-system' level, they are brought about by changes within constituent regulatory elements of that system (Kitano, 2004;Veraart et al, 2012). It has been put forward that disease manifests from a gradual deterioration of health (which is frequently reversible), through relatively sudden critical transitions that demark a definitive disease onset and are frequently irreversible (Chen, Liu, Li, & Chen, 2016;Li, Jin, Lei, Pan, & Zou, 2015).…”

Two Novel Methods for Extracting Synchronously Fluctuated Genes

“…The development of methods for transcriptome data analysis is one of the important research topics in the field of bioinformatics. In this study, I focus on the extraction of synchronously fluctuated genes (SFGs), which were also called dynamical network biomarkers in previous studies [1], [2], [3], [4], [5], [6], [7], [8], [9]. SFGs are similar to but different from differentially expressed genes (DEGs).…”

“…SFGs have been investigated by several research groups [1], [2], [3], [4], [5], [6], [7], [8], [9] based on an expectation that they are useful for characterizing the predisease state [1], that is, a state before the disease onset with lowered stability and abnormal homeostasis. The characterization of the predisease state may lead to rational and efficient disease prevention.…”

scPADGRN: A preconditioned ADMM approach for reconstructing dynamic gene regulatory network using single-cell RNA sequencing data

“…where the first term evaluates the precision of problem (5), the second term is the 154 L 1 -norm of the dynamic network to guarantee the sparsity of the network, and the third 155 term imposes the continuity assumption on the dynamic network states at consecutive 156 time points. Both sparsity and continuity need to be considered in biological 157 networks [3]. The parameters α and β are tuning parameters that control the penalties 158 for sparsity and continuity, respectively.…”

“…Bayesian information criterion (BIC) can be used to optimize the parameters α and 214 β [3]. Let L * denote the objective function of optimization problem (7).…”