1.  How are Predictive Models produced?

Models are typically produced from laboratory data where a food or microbiological broth is inoculated with a single bacterial species or a mixture of bacteria. If the researcher would like to model the effects of intrinsic factors of a food/broth, then a range of these factors are added to the sample, such as by adding various levels of salt, organic acid, or preservative. Next, the samples are incubated at various temperatures and the bacterial levels are measured at predetermined time intervals. After the experiments, the data are organized and the a curve is fitted to the data. This is termed Primary Modeling. In the case of growth, parameters for the lag time (if present), growth rate (or generation time) and the maximum population density are calculated. In the case of bacterial inactivation, the lag time (if present) and inactivation rate are calculated.  In Secondary Modeling, the change in a specific parameter is modeled as a function of a change in the environment (e.g., temperature, salt/water activity, pH) is modeled. The variation in predictions (upper and lower confidence levels) are also calculated. Finally, all of these data are developed into a model interface, like the PMP, so that users can easily make predictions of bacterial behavior. For more information on model development, go to the Toolbar, select References and click on “Model Development.”

2.  How can I find out more about how a specific PMP model was produced?

While the model window is open, the reference(s) or source of the information can be found in the “Source and/or Related Publications” window.