Using Multi-objective Optimization to Identify Dynamical Network Biomarkers as Early-warning Signals of Complex Diseases

Anticipating regime shifts by mixing early warning signals from different nodes

Real systems showing regime shifts, such as ecosystems, are often composed of many dynamical elements interacting on a network. Various early warning signals have been proposed for anticipating regime shifts from observed data. However, it is unclear how one should combine early warning signals from different nodes for better performance. Based on theory of stochastic differential equations, we propose a method to optimize the node set from which to construct an early warning signal. The proposed method takes into account that uncertainty as well as the magnitude of the signal affects its predictive performance, that a large magnitude or small uncertainty of the signal in one situation does not imply the signal's high performance, and that combining early warning signals from different nodes is often but not always beneficial. The method performs well particularly when different nodes are subjected to different amounts of dynamical noise and stress.

APELIN-13 AMELIORATES LPS-INDUCED ENDOTHELIAL-TO-MESENCHYMAL TRANSITION AND POST-ACUTE LUNG INJURY PULMONARY FIBROSIS BY SUPPRESSING TRANSFORMING GROWTH FACTOR-Β1 SIGNALING

ABSTRACT

The pathophysiology of acute respiratory distress syndrome (ARDS) involves cytokine storms, alveolar-capillary barrier destruction, and fibrotic progression. Pulmonary interstitial fibrosis is an important factor affecting the prognosis of ARDS patients. Endothelial-to-mesenchymal transition (EndMT) plays an important role in the development of fibrotic diseases, and the occurrence of EndMT has been observed in experimental models of LPS-induced acute lung injury (ALI). Apelin is an endogenous active polypeptide that plays an important role in maintaining endothelial cell homeostasis and inhibiting fibrotic progression in various diseases. However, whether apelin attenuates EndMT in ALI and post-ALI pulmonary fibrosis remains unclear. We analyzed the serum levels of apelin-13 in patients with sepsis-associated ARDS to examine its possible clinical value. A murine model of LPS-induced pulmonary fibrosis and an LPS-challenged endothelial cell injury model were used to analyze the protective effect and underlying mechanism of apelin-13. Mice were treated with apelin-13 by i.p. injection, and human pulmonary microvascular endothelial cells were incubated with apelin-13 in vitro . We found that the circulating apelin-13 levels were significantly elevated in sepsis-associated ARDS patients compared with healthy controls. Our study also confirmed that LPS induced EndMT progression and pulmonary fibrosis, which were characterized by decreased CD31 expression and increased α-smooth muscle actin expression and collagen deposition. LPS also stimulated the production of transforming growth factor β1 and activated the Smad signaling pathway. However, apelin-13 treatment significantly attenuated these changes. Our findings suggest that apelin-13 may be a novel biomarker in patients with sepsis-associated ARDS. These results demonstrate that apelin-13 ameliorates LPS-induced EndMT and post-ALI pulmonary fibrosis by suppressing transforming growth factor β1 signaling.

Plasma apelin and vascular endothelial growth factor levels in preterm infants: relationship to neonatal respiratory distress syndrome

AIM

The study aimed to determine the association between cord plasma levels of apelin and vascular endothelial growth factor (VEGF) with respiratory distress syndrome (RDS) in preterm infants.

METHODS

This case-control study included 120 preterm infants admitted to the neonatal intensive care unit of our hospital between January 2019 and January 2020. The infants were divided into RDS (n = 60) and non-RDS groups (n = 60). The cord plasma apelin and VEGF levels, perinatal characteristics, and neonatal complications were compared between the two groups.

RESULTS

The plasma apelin levels in the RDS group were significantly higher than in the non-RDS group (158.9 ± 24.8 vs. 125.2 ± 18.2 pg/mL, respectively), whereas VEGF levels in the non-RDS group were significantly higher than in the RDS group (187.4 ± 28.5 vs. 245.1 ± 44.8 pg/mL, respectively) (both p < .001). Infants with more severe RDS had higher plasma apelin levels and lower plasma VEGF levels. In the receiver operating characteristic curve analysis for the prediction of RDS, a cutoff of 148.4 pg/mL for apelin level yielded a sensitivity of 63.3% and a specificity of 95.0%, whereas a cutoff of 214.2 pg/mL for VEGF level showed a sensitivity of 86.7% and a specificity of 75.0%. Apelin levels were negatively correlated with VEGF levels in infants with RDS (r = 0.84, p < .001).

CONCLUSION

Differences in cord plasma apelin and VEGF levels may aid in the early diagnosis and treatment of RDS in preterm infants.

Apelin/APJ system: an emerging therapeutic target for respiratory diseases

Apelin is an endogenous ligand of G protein-coupled receptor APJ. It is extensively expressed in many tissues such as heart, liver, and kidney, especially in lung tissue. A growing body of evidence suggests that apelin/APJ system is closely related to the development of respiratory diseases. Therefore, in this review, we focus on the role of apelin/APJ system in respiratory diseases, including pulmonary arterial hypertension (PAH), pulmonary embolism (PE), acute lung injury (ALI)/acute respiratory distress syndrome (ARDS), obstructive sleep apnoea syndrome (OSAS), non-small cell lung cancer (NSCLC), pulmonary edema, asthma, and chronic obstructive pulmonary diseases. In detail, apelin/APJ system attenuates PAH by activating AMPK-KLF2-eNOS-NO signaling and miR424/503-FGF axis. Also, apelin protects against ALI/ARDS by reducing mitochondrial ROS-triggered oxidative damage, mitochondria apoptosis, and inflammatory responses induced by the activation of NF-κB and NLRP3 inflammasome. Apelin/APJ system also prevents the occurrence of pulmonary edema via activating AKT-NOS3-NO pathway. Moreover, apelin/APJ system accelerates NSCLC cells' proliferation and migration via triggering ERK1/2-cyclin D1 and PAK1-cofilin signaling, respectively. Additionally, apelin/APJ system may act as a predictor in the development of OSAS and PE. Considering the pleiotropic actions of apelin/APJ system, targeting apelin/APJ system may be a potent therapeutic avenue for respiratory diseases.

Identifying Early Warning Signals for the Sudden Transition from Mild to Severe Tobacco Etch Disease by Dynamical Network Biomarkers

Complex systems exhibit critical thresholds at which they transition among alternative phases. Complex systems theory has been applied to analyze disease progression, distinguishing three stages along progression: (i) a normal noninfected state; (ii) a predisease state, in which the host is infected and responds and therapeutic interventions could still be effective; and (iii) an irreversible state, where the system is seriously threatened. The dynamical network biomarker (DNB) theory sought for early warnings of the transition from health to disease. Such DNBs might range from individual genes to complex structures in transcriptional regulatory or protein-protein interaction networks. Here, we revisit transcriptomic data obtained during infection of tobacco plants with tobacco etch potyvirus to identify DNBs signaling the transition from mild/reversible to severe/irreversible disease. We identified genes showing a sudden transition in expression along disease categories. Some of these genes cluster in modules that show the properties of DNBs. These modules contain both genes known to be involved in response to pathogens (e.g., ADH2, CYP19, ERF1, KAB1, LAP1, MBF1C, MYB58, PR1, or TPS5) and other genes not previously related to biotic stress responses (e.g., ABCI6, BBX21, NAP1, OSM34, or ZPN1).

Detecting the stable point of therapeutic effect of chronic myeloid leukemia based on dynamic network biomarkers

Background

Most researches of chronic myeloid leukemia (CML) are currently focused on the treatment methods, while there are relatively few researches on the progress of patients’ condition after drug treatment. Traditional biomarkers of disease can only distinguish normal state from disease state, and cannot recognize the pre-stable state after drug treatment.

Results

A therapeutic effect recognition strategy based on dynamic network biomarkers (DNB) is provided for CML patients’ gene expression data. With the DNB criteria, the DNB with 250 genes is selected and the therapeutic effect index (TEI) is constructed for the detection of individual disease. The pre-stable state before the disease condition becomes stable is 1 month. Through functional analysis for the DNB, some genes are confirmed as key genes to affect the progress of CML patients’ condition.

Conclusions

The results provide a certain theoretical direction and theoretical basis for medical personnel in the treatment of CML patients, and find new therapeutic targets in the future. The biomarkers of CML can help patients to be treated promptly and minimize drug resistance, treatment failure and relapse, which reduce the mortality of CML significantly.

Electronic supplementary material

The online version of this article (10.1186/s12859-019-2738-0) contains supplementary material, which is available to authorized users.

Sparse Pathway-Induced Dynamic Network Biomarker Discovery for Early Warning Signal Detection in Complex Diseases

In many complex diseases, the transition process from the healthy stage to the catastrophic stage does not occur gradually. Recent studies indicate that the initiation and progression of such diseases are comprised of three steps including healthy stage, pre-disease stage, and disease stage. It has been demonstrated that a certain set of trajectories can be observed in the genetic signatures at the molecular level, which might be used to detect the pre-disease stage and to take necessary medical interventions. In this paper, we propose two optimization-based algorithms for extracting the dynamic network biomarkers responsible for catastrophic transition into the disease stage, and to open new horizons to reverse the disease progression at an early stage through pinpointing molecular signatures provided by high-throughput microarray data. The first algorithm relies on meta-heuristic intelligent search to characterize dynamic network biomarkers represented as a complete graph. The second algorithm induces sparsity on the adjacency matrix of the genes by taking into account the biological signaling and metabolic pathways, since not all the genes in the ineractome are biologically linked. Comprehensive numerical and meta-analytical experiments verify the effectiveness of the results of the proposed approaches in terms of network size, biological meaningfulness, and verifiability.

Optimal information networks: Application for data-driven integrated health in populations

Development of composite indicators for integrated health in populations typically relies on a priori assumptions rather than model-free, data-driven evidence. Traditional variable selection processes tend not to consider relatedness and redundancy among variables, instead considering only individual correlations. In addition, a unified method for assessing integrated health statuses of populations is lacking, making systematic comparison among populations impossible. We propose the use of maximum entropy networks (MENets) that use transfer entropy to assess interrelatedness among selected variables considered for inclusion in a composite indicator. We also define optimal information networks (OINs) that are scale-invariant MENets, which use the information in constructed networks for optimal decision-making. Health outcome data from multiple cities in the United States are applied to this method to create a systemic health indicator, representing integrated health in a city.

Several Indicators of Critical Transitions for Complex Diseases Based on Stochastic Analysis

Many complex diseases (chronic disease onset, development and differentiation, self-assembly, etc.) are reminiscent of phase transitions in a dynamical system: quantitative changes accumulate largely unnoticed until a critical threshold is reached, which causes abrupt qualitative changes of the system. Understanding such nonlinear behaviors is critical to dissect the multiple genetic/environmental factors that together shape the genetic and physiological landscape underlying basic biological functions and to identify the key driving molecules. Based on stochastic differential equation (SDE) model, we theoretically derive three statistical indicators, that is, coefficient of variation (CV), transformed Pearson's correlation coefficient (TPC), and transformed probability distribution (TPD), to identify critical transitions and detect the early-warning signals of the phase transition in complex diseases. To verify the effectiveness of these early-warning indexes, we use high-throughput data for three complex diseases, including influenza caused by either H3N2 or H1N1 and acute lung injury, to extract the dynamical network biomarkers (DNBs) responsible for catastrophic transition into the disease state from predisease state. The numerical results indicate that the derived indicators provide a data-based quantitative analysis for early-warning signals for critical transitions in complex diseases or other dynamical systems.

Modeling tumor progression via the comparison of stage-specific graphs

Can we use graph mining algorithms to find patterns in tumor molecular mechanisms? Can we model disease progression with multiple time-specific graph comparison algorithms? In this paper, we will focus on this area. Our main contributions are 1) we proposed the Temporal-Omics (Temp-O) workflow to model tumor progression in non-small cell lung cancer (NSCLC) using graph comparisons between multiple stage-specific graphs, and 2) we showed that temporal structures are meaningful in the tumor progression of NSCLC. Other identified temporal structures that were not highlighted in this paper may also be used to gain insights to possible novel mechanisms. Importantly, the Temp-O workflow is generic; while we applied it on NSCLC, it can be applied in other cancers and diseases. We used gene expression data from tumor samples across disease stages to model lung cancer progression, creating stage-specific tumor graphs. Validating our findings in independent datasets showed that differences in temporal network structures capture diverse mechanisms in NSCLC. Furthermore, results showed that structures are consistent and potentially biologically important as we observed that genes with similar protein names were captured in the same cliques for all cliques in all datasets. Importantly, the identified temporal structures are meaningful in the tumor progression of NSCLC as they agree with the molecular mechanism in the tumor progression or carcinogenesis of NSCLC. In particular, the identified major histocompatibility complex of class II temporal structures capture mechanisms concerning carcinogenesis; the proteasome temporal structures capture mechanisms that are in early or late stages of lung cancer; the ribosomal cliques capture the role of ribosome biosynthesis in cancer development and sustainment. Further, on a large independent dataset we validated that temporal network structures identified proteins that are prognostic for overall survival in NSCLC adenocarcinoma.

Identification of thyroxine-binding globulin as a candidate plasma marker of chronic obstructive pulmonary disease

Biomarkers for the management of chronic obstructive pulmonary disease (COPD) are limited. The aim of this study was to explore new plasma biomarkers in patients with COPD. Thyroxine-binding globulin (THBG) was initially identified by proteomics in a discovery panel and subsequently verified by enzyme-linked immunosorbent assay in another verification panel with a 1-year follow-up. THBG levels were elevated in patients with COPD (9.2±2.3 μg/mL) compared to those of the controls (6.6±2.0 μg/mL). Receiver operating characteristic curves suggested that THBG was able to slightly differentiate between patients with COPD and controls (area under the curve [AUC]: 0.814) and performed better if combined with fibrinogen (AUC: 0.858). THBG was more capable of distinguishing Global Initiative for Obstructive Lung Disease stages I–III and IV (AUC: 0.851) compared with fibrinogen (AUC 0.582). THBG levels were negatively associated with predicted percentage forced expiratory volume in 1 s and positively related to predicted percentage residual volume, RV/percentage total lung capacity, and percentage low-attenuation area. COPD patients with higher baseline THBG levels had a greater risk of acute exacerbation (AE) than those with lower THBG levels (P=0.014, by Kaplan–Meier curve; hazard ratio: 4.229, by Cox proportional hazards model). In summary, THBG is a potential plasma biomarker of COPD and can assist in the management of stable stage and AEs in COPD patients.