Computational modeling of selenoprotein deficiencies: a transformative avenue in selenoprotein-related psychiatric disorders

To the Editor, Selenoproteins are a distinct class of proteins that incorporate selenium (Se) in the form of the amino acid selenocysteine and are essential for a wide range of physiological processes, including immune function, DNA synthesis, redox signaling, and selenium transport. Selenoprotein P (SELENOP) functions as the main selenium transport protein in mammals and maintains systemic and cerebral selenium homeostasis. Although individual selenoproteins exert a variety of biological effects, their biosynthesis depends on a shared translational machinery and is highly dependent on dietary selenium availability. Clinical observations have increasingly implicated SELENOP in mood regulation, with lower circulating concentrations being associated with greater severity of depressive and anxiety symptoms, suggesting a potential contribution to affective functioning[1]. Historically regarded as a toxic trace element, selenium was not initially recognized for its essential biological roles; however, seminal experimental studies in 1957 demonstrated that selenium, in conjunction with vitamin E, is necessary for the prevention of liver necrosis[2]. Subsequent research established that selenoproteins exert their beneficial effects primarily through redox-regulatory mechanisms, including the neutralization of reactive oxygen species and protection against oxidative damage. Notably, several selenoproteins are highly expressed in the central nervous system, and accumulating evidence indicates their importance in normal brain development and neuronal function[3]. Selenoprotein deficiency has been associated with a range of psychiatric manifestations, including depression, anxiety, and cognitive dysfunction. Given the growing prevalence of mood disorders and the absence of a single unifying pathophysiological hypothesis, oxidative stress – an established feature of depression – has been proposed as a key mechanistic link between selenium deficiency and affective disorders[4]. Correction of selenoprotein deficiency is commonly approached through dietary selenium supplementation; however, understanding, predicting, and mitigating selenium-related dysfunction increasingly relies on computational modeling. In addition, selenium is essential for thyroid hormone metabolism through selenoprotein-dependent iodothyronine deiodinases, and selenium deficiency may disrupt hypothalamic–pituitary–thyroid axis activity, manifesting as mood disturbances, fatigue, cognitive slowing, and emotional dysregulation. In neurons particularly, where metabolic activity is high, effective antioxidant defense is critical to limit oxidative damage[5,6]. Several clinical and observational studies have reported that lower selenium levels are associated with increased severity of depressive symptoms[7]. Although the relationship is not fully understood, one possible explanation is that insufficient selenium weakens antioxidant systems, leading to increased oxidative stress. Oxidative stress has repeatedly been implicated in the pathophysiology of mood disorders and may influence neurotransmitter regulation, neuroinflammation, and neuronal survival[5,7,8]. At the cellular level, reduced selenoprotein activity may impair mitochondrial function and increase neuronal susceptibility to injury[6,8]. Mitochondrial dysfunction and increased oxidative stress can interfere with normal synaptic signaling and plasticity, both of which are important for maintaining stable mood and cognitive performance[8]. When these processes are disturbed, the impact is not confined to isolated cells. Gradually, such cellular-level changes may begin to affect wider neural networks. Evidence from both animal and human studies suggests that prolonged selenium deficiency and altered deiodinase function can significantly influence neurobehavioral outcomes[9,10]. Computational modeling has played a major role, giving substantial contributions to basic neuroscience and facilitating several advances in understanding brain functionality. Such approaches are increasingly applied in the study of psychiatric and neurological disorders, where they offer a formal framework for investigating complex neurobiological processes. Although computational models vary significantly in structure and purpose due to the heterogeneity of brain disorders, their overarching objective is to clarify the mechanisms linking neural activity to mental processes and clinical symptoms. A major strength of computational modeling lies in its ability to bridge multiple levels of explanation, from molecular and cellular neurobiology to cognitive function and the subjective experience of psychological distress. Computational approaches prove to be particularly effective in examining the mechanisms underlying brain disorders, as exemplified by studies of high-level visual processing[11]. In visual neuroscience, computational frameworks derived from information theory and predictive coding have successfully explained receptive field properties and early sensory processing, illustrating the explanatory power of computational modeling. Within this context, emerging evidence suggests that a deficiency of selenoprotein P results in reduced selenium availability in the brain, leading to impaired synaptic plasticity and alterations in cognitive and emotional regulation, further supporting the relevance of computational approaches for elucidating molecular-to-circuit-level contributions to psychiatric symptoms[8]. Despite the substantial contributions of computational frameworks to psychiatric research, these approaches are not without limitations. By necessity, computational models rely on simplified representations of complex biological systems, which may not fully capture the multilevel interactions among molecular, cellular, environmental, and developmental factors that influence selenoprotein function. Model outputs are inherently dependent on the quality, completeness, and accuracy of the input data, as well as on the assumptions underlying model construction; consequently, incomplete data or biologically implausible assumptions can constrain the validity of resulting inferences. In addition, the external validity of computational findings may be limited, as models developed under controlled or idealized conditions do not always generalize across heterogeneous populations or clinical contexts. Overreliance on in silico predictions, in the absence of rigorous experimental or clinical validation, may therefore lead to oversimplified mechanistic interpretations or suboptimal therapeutic strategies[10]. At present, there are no specific computational models that focus directly on selenoprotein deficiencies in psychiatric disorders. Future research should aim to develop mechanistic, multiscale models integrating genetic variation, selenium transport, oxidative stress pathways, and neural network function. In the longer term, integrating computational psychiatry with selenoprotein biology may help translate basic molecular findings into practical clinical tools. This approach has the potential to close existing gaps between biochemical abnormalities and psychiatric symptoms, leading to more targeted diagnostics and therapies. Through computational models, clinicians can better assess the risk, understand the possible treatment responses, and choose suitable nutritional or medical options.