U.S. flag

An official website of the United States government

Dot gov

Official websites use .gov
A .gov website belongs to an official government organization in the United States.


Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.


Main content area

Direct construction of predictive models for describing growth of Salmonella Enteritidis in liquid eggs – A one-step approach

Lihan Huang
Food control 2015 v.57 pp. 76-81
Salmonella Enteritidis, bacterial contamination, data collection, egg albumen, egg yolk, eggs, food contamination, growth models, logit analysis, microbial growth, prediction, rifampicin, temperature
The objective of this study was to develop a new approach using a one-step approach to directly construct predictive models for describing the growth of Salmonella Enteritidis (SE) in liquid egg white (LEW) and egg yolk (LEY). A five-strain cocktail of SE, induced to resist rifampicin at 100 mg/L, was used to inoculate LEW and LEY. Kinetic studies were conducted isothermally at different temperatures between 8 and 43 °C to generate growth curves at each temperature.This study first solved an inverse problem globally, using the growth curves to estimate the temperature-dependent kinetic parameters, and then applied the parameters to predict growth (a forward problem). Once the growth curves were generated, they were assembled and analyzed using nonlinear regression to determine kinetic parameters of both primary and secondary models in one step, with an objective to minimize the global residual sum of squares (RSS) for the entire data set. For growth in LEW, a three-parameter logistic model was used. For growth in LEY, the Huang model was used as the primary model. The Ratkowsky square-root model was used to evaluate the growth rates.The results showed that the one-step approach resulted in accurate estimation of the kinetic parameters that were used later to successfully predict the growth of SE in LEY and LEW. The estimated nominal minimum growth temperatures of SE were 7.4 °C and 9.9 °C, while the estimated maximum growth temperatures were 45.2 °C and 46.8 °C, respectively, in LEW and LEY. As a validation, the predictive models were tested with independent growth curves of SE in LEY and LEW at 37 °C. The root mean square error (RMSE) was only 0.36 and 0.28 log CFU/ml over a total scale of 8.4 and 7.8 log CFU/ml, respectively, for the growth models of SE in LEY and LEW, suggesting that the one-step approach can generate accurate models for predicting the growth of SE in LEY and LEW. The results from this study can be used to predict the growth of SE and evaluate the safety of LEY and LEW.