Bioinfo Chem

1, Introduction

Over the past few decades, biomedical research has been undergoing a subtle yet profound conceptual shift. What once relied heavily on reductionist frameworks—isolating single genes, proteins, or pathways—has gradually transitioned toward a more integrative, systems-oriented understanding of disease. This shift is not merely methodological; it reflects a growing recognition that complex diseases rarely arise from singular molecular events. Instead, they emerge from dynamic, nonlinear interactions within multilayered biological networks that span genomic, proteomic, and environmental dimensions (Kitano, 2002; Ideker et al., 2001).

Traditional genetic approaches, including linkage analysis and positional cloning, have undoubtedly been instrumental in identifying causative genes for Mendelian disorders. However, their explanatory power diminishes when applied to multifactorial diseases such as cancer, cardiovascular disorders, and neurodegenerative conditions. These diseases involve distributed perturbations across interconnected systems, where individual components contribute modest effects that collectively shape disease phenotypes (Xiong et al., 2005; Khatri et al., 2012). As such, there is an increasing need for frameworks capable of capturing the emergent properties of biological systems rather than focusing solely on their individual constituents.

Systems biology offers precisely this perspective. At its core, it emphasizes the integration of diverse “omics” datasets—genomics, transcriptomics, proteomics, and metabolomics—to reconstruct and analyze biological networks as cohesive entities (Kohl et al., 2010; Li, 2013). Rather than viewing genes or proteins in isolation, systems biology conceptualizes them as nodes within intricate interaction networks, where function arises from connectivity and context. This paradigm enables researchers to move beyond static representations of biological pathways toward dynamic models that reflect temporal behavior, feedback mechanisms, and system-level robustness.

One of the central challenges in systems biology lies in the integration of heterogeneous data sources. Single-omic datasets often suffer from inherent limitations, including noise, incomplete coverage, and context dependency. However, when combined, these datasets provide complementary insights that enhance the reliability and resolution of disease pathway models (Draghici et al., 2007; Rohart et al., 2017). For instance, integrating gene expression data with protein–protein interaction networks can illuminate the mechanistic routes through which genetic variations influence phenotypic outcomes. Such integrative approaches allow researchers to trace the progression from molecular perturbations to system-wide dysfunction, thereby offering a more comprehensive understanding of disease pathogenesis.

Modeling disease pathways within this framework involves several interconnected steps. First, the structural organization of the system must be defined, including the identification of key molecular components and their interactions. Next, system dynamics are analyzed to understand how these interactions evolve over time under varying conditions. Finally, computational and mathematical models—such as Bayesian networks, differential equations, and structural equation modeling—are employed to simulate system behavior and predict responses to perturbations (Kreeger & Lauffenburger, 2010; Edwards et al., 2011). These models are particularly valuable for capturing feedback loops and nonlinear dynamics that are difficult to infer using conventional statistical methods.

A defining feature of biological systems, and a critical consideration in disease modeling, is robustness. Biological networks are inherently resilient, maintaining functionality despite environmental fluctuations and internal disturbances. While this robustness is essential for normal physiological processes, it can also be co-opted by pathological systems. For example, cancer cells often exploit network robustness to sustain uncontrolled proliferation and resist therapeutic interventions (Kreeger & Lauffenburger, 2010). Systems biology approaches aim to identify critical nodes or “control points” within these networks—points where targeted perturbations can disrupt pathological stability and restore normal function.

The practical applications of systems biology are increasingly evident in the emerging fields of systems medicine and systems epidemiology. In pulmonary diseases such as asthma, integrative models incorporating genomic and proteomic data have been used to identify disease subtypes and predict treatment responses (Auffray et al., 2010). Similarly, in neurodegenerative diseases like Parkinson’s disease, combining genome-wide association studies with transcriptomic data has revealed consensus pathways implicated in disease progression, including axonal guidance and cellular adhesion processes (Edwards et al., 2011). In prion diseases, longitudinal analyses of gene expression have enabled the identification of core molecular signatures that drive neurodegeneration (Hwang et al., 2009). These examples illustrate how systems-level approaches can bridge the gap between molecular data and clinical insight.

Despite these advances, several challenges continue to constrain the full realization of systems biology in disease modeling. The sheer scale and complexity of biological networks demand sophisticated computational tools and substantial processing power. Moreover, discrepancies between in vitro findings and in vivo validation remain a persistent issue, highlighting the need for more physiologically relevant models. Data standardization and annotation also present significant hurdles, particularly in integrating environmental variables that influence disease pathways (Khatri et al., 2012). Addressing these challenges is essential for translating systems biology from a primarily exploratory discipline into a robust framework for precision medicine.

Within this evolving landscape, several critical questions naturally emerge, guiding both current inquiry and future investigation. A key consideration is whether the integration of multi-omics datasets meaningfully improves the accuracy of identifying causative disease pathways compared to traditional single-layer analyses. Closely related to this is the question of how effectively dynamic systems models can predict therapeutic outcomes, including unintended systemic effects, prior to clinical intervention. There is also growing interest in determining whether conserved network motifs exist across diverse disease categories, potentially serving as universal biomarkers or therapeutic targets. Furthermore, the influence of biological noise and environmental variability on the stability and predictive performance of disease pathway models remains an open and pressing challenge.

In light of these considerations, this narrative review is structured with several interconnected objectives. It seeks to evaluate the effectiveness of current computational frameworks—particularly Bayesian networks, differential equation models, and structural equation approaches—in capturing the complexity of biological interactions. It also aims to assess the role of multi-omics integration in enhancing the resolution and predictive capacity of disease pathway modeling. Beyond methodological evaluation, the review explores the translational potential of systems biology in identifying novel therapeutic targets for complex diseases characterized by high network robustness. Finally, it critically examines how environmental context and stochastic variability influence the reliability and applicability of systems-based disease models, thereby shaping their future role in precision medicine.

2. Methodology

2.1 Study Design and Review Framework

This study was conducted as a narrative review, aiming to synthesize and critically interpret existing literature on systems biology approaches to disease pathway modeling. Unlike systematic reviews that follow rigid inclusion–exclusion criteria, this narrative approach allows for a more flexible and integrative exploration of conceptual developments, computational frameworks, and translational applications within the field. The review was guided by an interpretive framework emphasizing network-based disease understanding, multi-omics integration, and computational modeling strategies (Kitano, 2002; Kohl et al., 2010). The objective was not only to summarize findings but also to identify emerging patterns, conceptual shifts, and methodological gaps across studies.

2.2 Literature Identification and Source Selection

Relevant literature was identified through a targeted exploration of peer-reviewed articles focusing on systems biology, computational modeling, and disease pathway analysis. Priority was given to seminal and highly cited works that have shaped the theoretical and methodological foundations of the field, including studies on network biology, pathway analysis, and integrative omics approaches (Ideker et al., 2001; Khatri et al., 2012). Additional studies were selected to represent advancements in disease-specific applications, such as neurodegenerative and pulmonary diseases (Auffray et al., 2010; Edwards et al., 2011).

The selection process emphasized conceptual relevance rather than exhaustive coverage. Studies were included if they contributed to at least one of the following domains: (i) systems-level modeling of biological processes, (ii) integration of heterogeneous omics datasets, or (iii) computational simulation of disease mechanisms. Foundational works on robustness and network theory were also incorporated to contextualize biological system behavior (Csete & Doyle, 2002).

2.3 Data Extraction and Thematic Synthesis

Rather than extracting quantitative data, this review employed a thematic synthesis approach. Key concepts, methodologies, and findings from selected studies were systematically categorized into thematic domains, including: network reconstruction, pathway analysis, system dynamics, and translational applications. This approach enabled the identification of recurring methodological patterns, such as the use of Bayesian networks, differential equation models, and pathway impact analysis (Draghici et al., 2007; Xiong et al., 2005).

The synthesis process also involved comparing methodological frameworks across studies to highlight differences in modeling strategies and analytical depth. For example, pathway-level analyses were contrasted with network-based approaches to assess their relative ability to capture system complexity and disease mechanisms (Khatri et al., 2012).

2.4 Integration of Computational Tools and Modeling Approaches

A central component of this review involved evaluating computational tools and modeling frameworks used in systems biology. Studies employing software platforms such as network visualization tools, pathway enrichment systems, and multi-omics integration frameworks were examined to understand their analytical capabilities and limitations (Wheelock et al., 2009; Glaab, 2018).

Special attention was given to integrative modeling approaches that combine multiple data layers, as these are critical for reconstructing disease pathways with higher resolution (Rohart et al., 2017). Additionally, simulation-based studies—including organism-level modeling efforts such as OpenWorm—were reviewed to assess the feasibility of predictive biological modeling (Sarma et al., 2018).

2.5 Critical Evaluation and Limitations of the Approach

The narrative methodology inherently involves subjective interpretation, which may introduce selection bias. However, this limitation was mitigated by grounding the analysis in well-established and widely cited literature. Furthermore, the review prioritizes conceptual coherence over quantitative aggregation, meaning that findings should be interpreted as analytical insights rather than statistical generalizations. Despite these limitations, the chosen methodology provides a comprehensive and flexible framework for understanding the evolving landscape of systems biology in disease modeling, allowing for the integration of diverse perspectives and the identification of future research directions (Kreeger & Lauffenburger, 2010; Li, 2013).

3. Integrating Computational Biology and Systems Analysis for Disease Pathway Modeling

3.1 From Reductionism to Network-Centric Thinking in Disease Biology

There is, perhaps, a quiet but undeniable shift unfolding in how we understand biological systems. For much of modern science, reductionism guided discovery—breaking complex systems into genes, proteins, and metabolites, with the expectation that understanding each piece would eventually explain the whole. This approach achieved remarkable successes, particularly in identifying causative genes for Mendelian disorders (Ideker et al., 2001). Yet, as research increasingly turned toward multifactorial diseases—cancer, cardiovascular disorders, neurodegeneration—it became clear that this framework, while powerful, was incomplete. These diseases do not arise from isolated defects but from dense webs of molecular interactions, shaped by feedback loops, environmental influences, and stochastic variability (Xiong et al., 2005; Auffray et al., 2010).

Systems biology emerged as a response to this limitation, offering a more integrative lens through which biological complexity could be examined. Rather than focusing on individual components, it considers how networks behave—how relationships, rather than isolated elements, define function and dysfunction (Kitano, 2002). This shift reframes biological inquiry itself, moving from static descriptions of pathways toward dynamic interpretations of systems. It also raises important questions: can understanding network behavior reveal disease mechanisms that reductionist approaches overlook? And perhaps more importantly, can such an approach transform how we predict, diagnose, and treat disease?

3.2 Computational Biology and Systems Analysis: Tools for Understanding Complexity

At the heart of this transformation lies the synergy between computational biology and systems analysis. Computational biology provides the technical capacity to handle the overwhelming scale of modern biological data. Through knowledge discovery techniques—ranging from bioinformatics pipelines to machine learning—researchers can extract patterns from high-throughput omics datasets, uncovering relationships that would otherwise remain hidden (Kitano, 2002; Xiong et al., 2005). Yet, identifying patterns is only the beginning. Simulation-based approaches extend this process, enabling researchers to test hypotheses in silico, model perturbations, and explore system behavior before experimental validation (Ideker et al., 2001). This iterative loop—data, model, simulation, refinement—has become a defining feature of systems biology.

Systems analysis complements this by providing a conceptual and mathematical framework to interpret these patterns. It emphasizes system structure, mapping the architecture of molecular interactions; system dynamics, understanding how these interactions evolve over time; and control mechanisms, identifying points where intervention can restore or disrupt function (Kitano, 2002; Kohl et al., 2010). Mathematical tools—including differential equations, Bayesian networks, and structural equation modeling—allow researchers to capture nonlinear dependencies, feedback loops, and probabilistic relationships inherent in biological systems (Ideker et al., 2001; Xiong et al., 2005). Increasingly, a “middle-out” modeling strategy is favored, balancing molecular detail with system-level insight, thereby bridging the gap between reductionist and holistic approaches (Kohl et al., 2010).

The increasing sophistication of systems biology has also been shaped by foundational work in network robustness, evolutionary modeling, and computational pattern analysis. Early studies on biochemical robustness demonstrated that even simple regulatory networks can maintain stability under perturbations, highlighting resilience as an intrinsic system property rather than an emergent accident (Barkai & Leibler, 1997). Building on this, advances in probabilistic and evolutionary modeling frameworks—such as Bayesian phylogenetics—have expanded the capacity to infer dynamic biological processes across temporal and genetic scales (Bouckaert et al., 2019). At the network level, classification approaches integrating molecular interaction data have further illustrated how disease phenotypes, particularly metastasis, can be predicted through system-wide patterns rather than isolated markers (Chuang et al., 2007). Moreover, computational models of developmental patterning have provided critical insights into how spatial and temporal coordination emerges from interacting biological components, reinforcing the importance of multiscale modeling in understanding complex disease pathways (Morelli et al., 2012).

3.3 Robustness, Disease Networks, and Translational Insights

One of the most intriguing aspects of biological systems is their robustness—their ability to maintain stability despite internal noise and external perturbations. This resilience is not accidental but emerges from network architecture, including redundancy and feedback control mechanisms (Kitano, 2002; Csete & Doyle, 2002). However, in disease contexts, robustness becomes paradoxical. Pathological systems, particularly cancer, often exploit these same principles to sustain growth and resist therapeutic interventions (Kreeger & Lauffenburger, 2010). Systems biology, therefore, seeks not only to understand robustness but to identify its vulnerabilities—the critical nodes or “Achilles’ heels” within disease networks that can be targeted for intervention.

These insights are increasingly shaping clinical research. In pulmonary diseases such as asthma, systems-level analyses have revealed why targeting single inflammatory mediators is often ineffective, highlighting instead the need for multi-target therapeutic strategies (Auffray et al., 2010). In neurodegenerative diseases, integrative analyses have identified shared pathways—such as axonal guidance and focal adhesion—that underlie disease progression (Edwards et al., 2011). Similarly, longitudinal studies in prion disease have uncovered core gene expression signatures that define neurodegenerative trajectories (Hwang et al., 2009). Together, these examples illustrate how systems biology can move beyond descriptive models toward actionable insights, bridging molecular mechanisms with clinical outcomes.

3.4 Future Directions, Open Questions, and Emerging Challenges

Despite its promise, systems biology remains, in some ways, a field still finding its footing. The integration of heterogeneous datasets continues to pose challenges, with issues of data standardization, annotation, and interoperability limiting the full potential of multi-omics approaches (Wheelock et al., 2009). Moreover, translating computational predictions into clinically validated outcomes remains a significant hurdle, underscoring the persistent gap between in silico modeling and in vivo reality (Kreeger & Lauffenburger, 2010). As models grow increasingly complex, there is also a growing need for higher-level computational frameworks capable of managing this complexity without sacrificing interpretability (Purnick & Weiss, 2009; Sarma et al., 2018). Within this evolving landscape, several critical questions emerge. To what extent can multi-omics integration truly enhance the accuracy of disease pathway identification? How reliably can computational models predict therapeutic outcomes before clinical testing? Are there conserved network motifs across diseases that could serve as universal biomarkers? And how do environmental variability and biological noise influence model stability? Addressing these questions is central to advancing the field. Projects such as OpenWorm hint at a future where fully integrative, multiscale simulations may enable virtual experimentation at the organism level (Sarma et al., 2018). While such ambitions remain aspirational, they reflect a broader shift toward predictive, systems-driven medicine—one that may ultimately redefine how we understand and intervene in human disease.

4. Systems Biology and the Reconfiguration of Disease Understanding

4.1 Beyond Expression: Reinterpreting Control Within Disease Networks

One of the more subtle—but profoundly important—insights emerging from systems biology is that biological importance is not always aligned with what is most visibly altered. For years, much of molecular research has relied on identifying genes that show the greatest differential expression between healthy and diseased states. It seems intuitive: the louder the signal, the more important it must be. And yet, when examined through the lens of systems analysis, this assumption begins to unravel. Structural equation modeling, for instance, reveals that genes with only modest expression changes can exert disproportionately large regulatory influence within a network. In fibrotic diseases such as systemic sclerosis, genes like COL11A1 transition from relatively minor participants to dominant regulators, not because their expression skyrockets, but because their position within the network—and thus their control over downstream interactions—shifts dramatically (Xiong et al., 2005). The differential expression and regulatory strength of key genes involved in TGF-β signaling pathways are summarized in Table 1, highlighting their functional roles in fibrosis and extracellular matrix remodeling.

This distinction between expression and regulatory effect invites a reconsideration of how we define biological relevance. A gene may be highly expressed yet functionally peripheral, while another, less conspicuous gene may act as a critical hub, orchestrating system-wide responses. From a therapeutic perspective, this difference is not trivial. Targeting the most “visible” gene may yield limited results if that gene does not govern network behavior. Systems biology, therefore, encourages a shift toward identifying control nodes—points within the network where perturbation can propagate meaningful change. It is, in a sense, a move from observing the symptoms of disease to understanding its underlying circuitry.

4.2 Model Organisms and the Emergence of Predictive Biological Simulation

The effort to understand such circuitry has long relied on model organisms, which serve as simplified representations of biological complexity. There is a certain intellectual leap involved in assuming that principles derived from bacteria, flies, or worms can inform human disease, yet this assumption has proven remarkably productive. Early computational models of Escherichia coli and bacteriophage systems demonstrated that even relatively simple biological networks could exhibit rich, dynamic behaviors when analyzed through mathematical frameworks (Kitano, 2002; Ideker et al., 2001). The diversity of computational and systems biology approaches across different model organisms is outlined in Table 2, illustrating the range of modeling frameworks and simulation strategies employed in systems-level research.  These foundational studies laid the groundwork for extending systems approaches to multicellular organisms, where developmental processes revealed how robustness, feedback, and regulatory logic shape biological outcomes.

What is perhaps most striking is how far this trajectory has progressed. Projects like OpenWorm attempt to simulate the entire organismal behavior of Caenorhabditis elegans, integrating neuronal activity, muscle dynamics, and environmental interactions into a unified computational framework (Sarma et al., 2018). Such efforts signal a transition from descriptive biology to predictive simulation. The idea that we might one day construct “digital twins” of biological systems—models that can forecast disease progression or therapeutic response—is no longer purely speculative. Still, it raises important questions about scale, data quality, and interpretability. Can the complexity of human biology truly be captured in silico? And if so, how do we ensure that these models remain grounded in biological reality rather than computational abstraction?

4.3 Consensus Pathways, Robustness, and the Logic of Disease Progression

If systems biology offers one clear advantage, it is its ability to cut through the noise inherent in high-throughput data. The integration of multiple omics layers allows researchers to identify consensus pathways—those that consistently emerge across independent datasets and methodologies. In neurodegenerative diseases such as Parkinson’s disease, this approach has revealed pathways like axon guidance and focal adhesion as central drivers of disease progression (Edwards et al., 2011). The strength of these findings lies not only in their statistical

Table 1. Gene Expression and Regulatory Strength in TGF-β Pathways. This table describes the regulatory dynamics and statistical profiles of genes involved in TGF-beta pathways, specifically within the context of Systemic Sclerosis (SSc) studies (Xiong et al., 2005). It highlights statistically significant changes in gene regulation and their biological roles in fibrosis and extracellular matrix remodeling.

Table 2. Systems Biology Approaches Across Model Organisms. This table synthesizes diverse computational approaches and modeling frameworks used to decode biological systems across various organisms (Ideker et al., 2001; Sarma et al., 2018). It integrates methodological diversity, simulation frameworks, and data sources used in systems biology research.

Table 3. Comparative Pathway Significance in Complex Diseases. This table presents pathway-level statistical significance derived from integrative omics analyses in Parkinson’s disease and metabolic stress conditions. This table provides a meta-analysis of over-represented pathways in Parkinson's Disease and Palmitate response, showing statistical significance across different "omic" analyses (Edwards et al., 2011; Draghici et al., 2007). It highlights multi-level evidence (GWAS, expression, meta-analysis) supporting pathway involvement in disease mechanisms.

significance but in their reproducibility across different data types, lending them a level of credibility that single-layer analyses often lack.

At the same time, advances in pathway analysis methods have refined how these pathways are interpreted. Traditional over-representation analysis treats all genes within a pathway as equal, but systems approaches recognize that position matters. Pathway impact analysis, for example, accounts for network topology, highlighting how disruptions in central nodes can have far greater consequences than changes in peripheral components (Draghici et al., 2007). This shift in perspective aligns closely with the concept of robustness—a defining feature of biological systems. Robustness allows systems to maintain function despite perturbations, yet in disease contexts, it can also enable pathological stability. Cancer exemplifies this paradox, where robust signaling networks support survival even under therapeutic pressure (Kreeger & Lauffenburger, 2010). Pathway-level statistical significance across complex diseases, including Parkinson’s disease and metabolic stress conditions, is presented in Table 3, highlighting multi-omics evidence supporting key disease-associated pathways.

Understanding this duality—robustness coupled with fragility—is central to systems medicine. While networks are designed to resist change, they inevitably contain points of vulnerability. Identifying these fragile nodes offers a strategic advantage, allowing interventions to disrupt disease processes with minimal collateral effects. Importantly, such targets are not always intuitive, underscoring the value of computational modeling in revealing hidden dependencies within complex networks (Csete & Doyle, 2002).

4.4 Translational Challenges and the Future of Systems-Driven Medicine

Despite its conceptual and methodological advances, systems biology faces a series of practical challenges that cannot be overlooked. Data heterogeneity remains a significant barrier, with variations in experimental design, measurement techniques, and annotation standards complicating efforts to integrate datasets (Wheelock et al., 2009). Moreover, the translation of computational predictions into clinically validated outcomes is far from straightforward. Models may perform well under controlled conditions yet struggle to capture the variability inherent in real-world biological systems (Kreeger & Lauffenburger, 2010).

The expanding computational toolbox—encompassing platforms such as Cytoscape for network visualization and mixOmics for multi-omics integration—has undoubtedly enhanced analytical capabilities (Rohart et al., 2017; Wheelock et al., 2009). However, as Kohl et al. (2010) aptly noted, these tools are aids to thought, not replacements for it. The interpretation of systems-level data still requires biological insight, careful validation, and, perhaps most importantly, interdisciplinary collaboration. An overview of widely used computational tools for systems biology analysis, including platforms for network visualization and multi-omics integration, is provided in Table 4.  The dialogue between computational scientists and experimental biologists remains essential, ensuring that models are both technically robust and biologically meaningful.

Looking forward, the trajectory of systems biology appears to be moving toward increasingly integrative and predictive frameworks. The ambition to simulate entire biological systems—whether organs or organisms—suggests a future where disease modeling is not only descriptive but anticipatory. Yet, this vision must be approached with caution. As models grow in complexity, maintaining interpretability and clinical relevance becomes increasingly challenging. The field must balance its pursuit of comprehensive modeling with the need for actionable insights.

In conclusion, systems biology represents more than a methodological evolution; it reflects a deeper shift in how we conceptualize disease. By moving beyond static, reductionist models toward dynamic, network-based interpretations, it offers a more nuanced understanding of biological systems—one that acknowledges both their complexity and their underlying logic. While challenges remain, the integration of computational biology and systems analysis has already begun to reshape biomedical research, laying the groundwork for a future in which disease is not merely treated, but anticipated and strategically controlled.

5. Limitations

This review, while comprehensive in scope, is inherently constrained by its narrative design. The selection of literature, although grounded in influential and widely cited studies, reflects a degree of interpretive bias that cannot be entirely eliminated. Unlike systematic reviews,

Table 4. Computational Tools for Systems Biology Analysis. This table lists widely used computational tools for systems biology, highlighting their analytical functions and visualization capabilities. This table outlines a selection of software tools used for data integration, network visualization, and pathway analysis (Wheelock et al., 2009; Glaab, 2018). It provides a comparative overview of software platforms used for network analysis, pathway enrichment, and multi-omics integration.

this approach does not employ formal inclusion–exclusion criteria or quantitative synthesis, which may limit reproducibility and introduce subjectivity in thematic emphasis.

Additionally, the rapidly evolving nature of systems biology means that certain emerging tools, datasets, and computational frameworks may not be fully represented. The review also relies heavily on published studies, which may themselves be subject to publication bias, particularly favoring successful or novel findings over negative results. Finally, while the discussion highlights translational potential, many of the computational models examined remain insufficiently validated in clinical contexts, limiting the immediate applicability of these approaches to real-world medical decision-making.

6. Conclusion

Systems biology represents a meaningful shift in how disease is conceptualized—moving from isolated components toward interconnected networks that evolve over time. By integrating multi-omics data with computational modeling, it offers a framework capable of capturing the complexity and dynamism of biological systems. Yet, its true value may lie not only in improved understanding but in its potential to anticipate disease behavior and guide intervention strategies. Bridging the gap between computational prediction and clinical application remains a central challenge, but one that, if addressed, could fundamentally transform precision medicine and therapeutic design.

Author Contributions

R.G. conceptualized the study, designed the review framework, and drafted the original manuscript. A.K. contributed to literature analysis, interpretation of findings, and critically reviewed and edited the manuscript for important intellectual content.  All authors read and approved the final version of the manuscript.

References