Systems biology as a tool to approach complexity in neuroscience

• ¹ New York University Abu Dhabi, United Arab Emirates

The environment exerts its influence on multiple levels of biological organization, from molecular and cellular phenotypes to physiological and behavioral traits. These influences are measurable and can be quantified to understand the architecture of physiological and pathophysiological phenotypes (Idaghdour & Awadalla, 2013)(Fig. 1A). Molecular processes are subject to the effects of genetic and external factors, which can affect cells, tissues, organs and systems. However, this linear perspective must be contextualized for the complexity of interactions transcending from genes to neural and cognitive systems and potentially affecting cognitive traits and brain architecture and causing diseases of the nervous system (Fig 1B). Recent research has taken advantage of the wealth of accessible genomic information and multiple omics levels to identify genetic factors governing these relationships. Paradoxically, these studies reveal that the complexity of these systems is even greater than previously appreciated warranting the need for further application of interdisciplinary approaches in neuroscience. One of the trends that has fueled this progression is meta-analyses of multiple datasets provided by large cohort studies such as CHARGE, COGENT, and the UK Biobank, allowing for the discovery of genetic factors with low effect sizes. Tests of association of single nucleotide polymorphisms (SNPs) with cognitive traits shed light on genes and pathways involved in the molecular mechanisms of cognition (Lee et al., 2018). These links also shortlist interactions or dependences, albeit lacking directionality, between the nodes shown in Fig 1B (Howard et al., 2019). The overlap of some of the pathways identified with those previously associated with Alzheimer’s disease (Jansen et al., 2019) highlight the intricacy of molecular mechanisms underlying normal and disease variation of the nervous system. A deeper analysis of the data presented by these gargantuan projects belie the limitations of current model systems in capturing the molecular and neuronal orchestra involved in cognition and neurodegeneration. While a reductionist approach yields useful clinical insight for testing novel therapeutics, it is often argued that these models do not accurately mimic their respective diseases due to differences in genomic and systemic structure that compound in complexity in downstream processes (Henstridge, Hyman, & Spires-Jones, 2019; Zhao & Bhattacharyya, 2018). These systemic differences and the pressing need for efficient translational research have pushed scientists to explore alternative models for specific diseases (Saleem & Kannan, 2018; Zhao & Bhattacharyya, 2018). The dependence of high-level cognitive function on brain morphology provides the added challenge of merging genetics and structural analysis methods. Recent studies are now combining neuroimaging and genetic techniques to identify the structure-specific SNPs and measure the heritability of brain structural features, including volumetric measures and cortical thickness (Elliott et al., 2018; Howard et al., 2019; Sudre et al., 2019). These findings are particularly relevant in pathological contexts; a theme of neurodegenerative diseases is the selective vulnerability of certain structures or cell types, and may help elucidate resilience or susceptibility/resistance pathways. Much like the sugar-laden frosting of a Krispy Kreme* donut, environmental influences on the genome and downstream omics add a layer of complexity to the relationships between the nodes shown in Fig. 1B. As an example, the microbiome has been associated with both healthy neurodevelopment and pathological processes. Consequently, various perspectives have been employed to explore this bidirectional nature, linking the microbiome to diffusion tensor imaging phenotypes of the brain (Ong et al., 2018), brain metabolism (Soto et al., 2018) and neuroinflammation (Wong et al., 2016). This global influence demonstrates the necessity of a systems perspective that combines molecular, cellular, physiological and behavioral information to help unravel the architecture of complex traits of interest in neuroscience. The particularly pervasive insight that is coming from the fields of genomics, molecular neuroscience and brain imaging is a renewed appreciation of the complexity of neural systems and associated phenotypes. We argue that understanding the basis of these systems and phenotypes requires the adoption of multidisciplinary approaches that embrace complexity of cognitive traits and neurodegenerative disease. Systems biology approaches enhanced our ability to integrate data, techniques and knowledge across various levels of biological organization and across disciplines thus providing a convenient framework to conceptualize and study such complex traits and diseases. *not a sponsor Figure 1: Relationships and interactions modulating brain function. A Simplified view of genetic and environmental factors affecting phenotype. Gene-environment interactions can act at any level of the genotype-phenotype axis. B These interactions transcend to neuronal systems modulating the relationship between cognitive traits, brain structure and architecture, and neurodegenerative diseases. The interactions and directionality between these nodes have yet to be fully resolved, but integrative approaches are beginning to fill the gap.