Introduction

Model organisms are a powerful resource for biomedical research.
Outbred populations, such as the Heterogeneous Stock (HS) rats, are well-suited for identifying shared genetic architecture between physiological traits.
Here, we provide a tool to predict fasting glucose levels using user input values for body weight and retroperitoneal fat pad mass.
Findings from this research area will be relevant for identifying risk factors for Type 2 diabetes and developing personalized medicine.

Background

Chitre et al. (2020) performed a genome-wide association study in over 3,000 outbred HS rats and identified genetic loci that influence several physiological traits.
The authors found that fasting glucose levels are genetically correlated with body weight and retroperitoneal fat pad mass.
Here, we aimed to use body weight and retroperitoneal fat pad mass as covariates to predict fasting glucose levels. A random forest prediction algorithm (caret package) was trained on the publicly available dataset.
In the app, users can move two sliders to change the values of the covariates and click a button to predict the resting glucose levels. The output includes the predicted glucose level printed as text and a 3d scatterplot with the predicted data point overlaid on the observed data.

GitHub Repo: https://github.com/epduffy/DevelopingDataProducts_CourseProject
Shiny App: http://epduff.shinyapps.io/RatPhysiology
Reference: Chitre, A.S., Polesskaya, O., Holl, K., Gao, J., Cheng, R., Bimschleger, H., Garcia Martinez, A., George, T., Gileta, A.F., Han, W., Horvath, A., Hughson, A., Ishiwari, K., King, C.P., Lamparelli, A., Versaggi, C.L., Martin, C., St. Pierre, C.L., Tripi, J.A., Wang, T., Chen, H., Flagel, S.B., Meyer, P., Richards, J., Robinson, T.E., Palmer, A.A. and Solberg Woods, L.C. (2020), Genome-Wide Association Study in 3,173 Outbred Rats Identifies Multiple Loci for Body Weight, Adiposity, and Fasting Glucose. Obesity, 28: 1964-1973. https://doi.org/10.1002/oby.22927