Toward the Next Generation of In Silico Modeling of Dynamic Host-Microbiota Interactions in the Skin

Abstract

Understanding how the skin microbiota contributes to skin health and disease requires knowledge of the dynamic interactions between the skin and its resident microbes. In silico modeling complements in vivo and in vitro experiments by enabling a systems-level understanding of dynamic skin-microbiota interactions. However, the number of published in silico skin microbiota models remains limited. This paper provides the first comprehensive exploration of in silico skin microbiota modeling. We identify current challenges, learn from leading experimental validation approaches adopted in in silico gut microbiota research, and propose ways to enhance the predictive power of in silico skin microbiota models.

Keywords: In silico modeling; skin disease; skin microbiome; time-course data; validation.

Figure 1

Comparison between the in silico skin and gut microbiota models reviewed. (a) Trends in the number of in silico gut (red) and skin (blue) models over the years. (b) Pain points (red text) in the “learn-build-predict-validate” cycle of in silico skin microbiota modeling. The box on the left summarizes the number of studies that validated model predictions, whereas the box on the right categorizes studies on the basis of how model parameters were derived, whether using guesstimates, literature data, or specifically collected empirical data.

Figure 2

Proposed approaches to collecting time-course experimental data for in silico skin microbiota modeling. (a) Building up to a systems-level model of the skin microbiota through submodels of microbe-microbe and skin-microbe interactions. (b, c) Time-course data that can be used to respectively build and validate in silico models of microbe-microbe and skin-microbe interactions.