Dose Response Modeling R Code

From QMRAwiki
Jump to: navigation, search

Dose Response Modeling

Author: Mark H. Weir Ph.D. - The Ohio State University

General overview

Classical dose response modelling is most typically an exercise in optimization of the current dose response models. For more in-depth information regarding the development of the exponential and beta Poisson dose response models can be found in Haas et al (2014) [1]. Weir (2016) outlines the optimization and bootstrapping methodology with illustrative examples within the context of environmental and clinical pathogens [2].

In essence the exponential dose response model is derived by assuming that the exposed dose to a host is Poisson distributed for the average of the bolus dose. This means that we assume the distribution of the average of the bolus dose is random in description. Then the probability that the pathogens will survive to initiate an infection is best described using a Binomial distribution. This means that either the pathogen survives or does not to initiate an infection. The beta Poisson dose response model is an expansion from the exponential dose response model. In this case rather than assuming the rate term in the binomial distribution is constant we allow it to vary based on a beta distribution. This allows for an increased level of realism in modeling the dose response relationship in the host.

Dose Response Modeling in R

R is an ideal environment to develop a dose response modeling script, primarily due to the open source nature of the platform. The R environment provides high quality powerful computational tools in an open source environment. The source code uploaded onto this QMRA Wiki allows users to have the computational power of 18 peer reviewed publications as well as all of the dose response models found here on the QMRA Wiki. The code that can be downloaded here is also the backbone of the Use Experimental Data portion of VizDR the Dose Response App.

Use of the code and app is detailed and outlined in Weir et al. (2016 in press)[3]. The manuscript published in Environmental Modelling and Software will highlight how the code works and how the associate app, VizDR, works as well.


There is no warranty to this code. The instructions for use within the code are simple in that the data is imported to R and then the script is sourced in R. Any damage from the use of the source code is at the risk of the user, however, there has never been an issue with using the base code alone as opposed to customization or alterations. Alterations that require collaborations can be conducted by contacting the primary author of the code Dr. Mark H. Weir.


Please navigate to the file location. Note that the file is a pdf written in knitr RMarkdown with the above description as well as the code. Code


  1. Haas, C.N., Rose, J.B. and Gerba, C.P. 2014. Quantitative Microbial Risk Assessment. Second. J. Wiley and Sons. Wiley Link
  2. Weir, M.H. 2016. “Dose-Response Modeling and Use: Challenges and Uncertainties in Environmental Exposure.” In Manual of Environmental Microbiology, 4th ed., 3.5.3–1 – 3.5.3–17. ASM Press. Link to ASM book page
  3. Weir, M.H., Mitchell, J., Flynn, W.B., Pope, J.M. 2016(in press). Development of a Microbial Dose Response Visualization and Modelling Application for QMRA Modelers and Educators. Environmental Modelling and Software