Laboratory of Food Microbiology and Food Preservation, Department of Food Safety and Food Quality, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium.
Int J Food Microbiol. 2009 Oct 31;135(2):83-9. doi: 10.1016/j.ijfoodmicro.2009.07.038. Epub 2009 Aug 8.
This research is an extension of previous work reported in Gysemans et al. [Gysemans, K.P.M., Bernaerts, K., Geeraerd, A.H., Vermeulen, A., Debevere, J., Devlieghere, F., Van Impe, J.F., 2007. Exploring the performance of logistic regression model types on growth/no growth data of Listeria monocytogenes. International Journal of Food Microbiology 114, 316-331.] in which the growth/no growth interface of Listeria monocytogenes was modelled as a function of water activity (a(w)), pH and undissociated acetic acid percentage (UAc). The major difference with the previous work is that in the present research the influence of the cell density (N) is also considered during the modelling process. New experimental data were therefore collected as a function of a wide range of cell densities up until the level of the individual cell. Prior to the development of model that incorporates N, the expected inadequacy of the high cell density growth/no growth model developed in Gysemans et al. (2007) on the new cell density dependent data was illustrated. Inadequacy of the model at lower cell densities was expected since the data showed a significant reduction of the growth probability as N decreased. For the development of a model that incorporates the effect of N, a square-root type logistic regression model was proposed and evaluated. The model predicts a strong influence of the cell density with an increase in the growth probability if the cell count increased. The onset of this increase is dependent on the intrinsic factors of the medium (pH, a(w), and acetic acid concentration). The model also suggests that it is unlikely that a larger population has a higher chance to start growing just because the chance on a strong cell is higher in a larger population. It seems that the bacteria influence each other's growth.
本研究是对 Gysemans 等人之前工作的扩展[Gysemans, K.P.M., Bernaerts, K., Geeraerd, A.H., Vermeulen, A., Debevere, J., Devlieghere, F., Van Impe, J.F., 2007. 探索逻辑回归模型类型在李斯特菌生长/不生长数据上的性能。国际食品微生物学杂志 114, 316-331.]，其中李斯特菌的生长/不生长界面被建模为水活度（a(w)）、pH 值和未解离乙酸百分比（UAc）的函数。与之前工作的主要区别在于，在本研究中，在建模过程中还考虑了细胞密度（N）的影响。因此，作为细胞密度广泛变化的函数，新的实验数据被收集，直到单细胞水平。在开发包含 N 的模型之前，说明了 Gysemans 等人（2007 年）开发的高细胞密度生长/不生长模型在新的细胞密度相关数据上的预期不适用性。由于数据显示随着 N 的降低，生长概率显著降低，预计模型在较低细胞密度下不适用。为了开发包含 N 影响的模型，提出并评估了平方根类型的逻辑回归模型。该模型预测细胞密度的强烈影响，如果细胞计数增加，生长概率会增加。这种增加的开始取决于培养基的内在因素（pH 值、a(w)和乙酸浓度）。该模型还表明，仅仅因为较大的群体中较强的细胞的机会更高，较大的群体不一定有更高的开始生长的机会。似乎细菌会相互影响彼此的生长。
