Modeling the Gut-Skin axis
The gut microbiome has a significant effect on every part of human physiology in ways that are only now making themselves understood. Intestinal dysbiosis (i.e. a state of microbial imbalance) has the potential to negatively impact gastrointestinal, immune, and neuroendocrine systems and has been linked to a multitude of disease states ranging from diabetes to mood disorders. It can also contribute to common skin disorders such as acne, psoriasis, eczema and rosacea [1]. Consequently, a bidirectional connection has been observed between the gut microbiome and the skin (i.e. the gut–skin axis). Ongoing clinical trials reveal that the gut–skin axis may also potentially contribute to autoimmune diseases of the skin [2].
Traditionally, clinical trials and statistical comparison between large groups with similar pathologies have been used for the diagnostic and therapeutic assessment of patients. However, there is a growing realization that patient groups are less uniform and more gradation is necessary for therapeutic intervention. Consequently, the identification of better treatments is difficult in the conventional empirical setup of clinical trials for gut mediated skin diseases.
We believe that there is an inherent need for patient-specific models that drive the development of techniques for creating personalized methods to guide pharmaceutical and diet therapies, as well as the development and deployment of synthetic probiotic devices.
Our research program seeks to develop models based on physiology and physics rather than on population statistics. This enables the discovery of underlying mechanisms linked to diagnostic information that would have otherwise remained concealed.
These models can provide useful predictions of individual patient treatment options by the adjustment of a well-chosen set of biomedically relevant parameters. One key challenge in our approach is the incorporation of complex multi-scale interactions across multi-organ sub-systems that contribute to system level behaviors. To make progress in model development, it is important to work alongside biologists and clinicians creating fundamentally new theoretical analysis tools, generating mechanistic hypothesis and creating quantitative models with close proximity to clinical and experimental observations.
With the integration of control theory, dynamical systems, and computational techniques we hope to reduce translational barriers regarding model personalization, speed, and practical use in a clinical environment to create a "virtual patient" capable of diagnosis and in silico optimization of treatment.
[1] J. Dou, M. R. Bennett, Biotechnol J. 13, e1700159–e1700159 (2018).
[2] W. E. Ruff, T. M. Greiling, M. A. Kriegel, Nature Reviews Microbiology. 18, 521–538 (2020).