Predicting the combined effect of Zataria multiflora essential oil, pH, and temperature on the growth of Staphylococcus aureus using artificial neural networks

Zataria multiflora is a plant belonging to the Laminaceae family that grows in Iran, Pakistan and Afghanistan. Although artificial neural network is a sophisticated novel method in predictive microbiology, no study has been fulfilled on Z. multiflora. Therefore, the present study is undertaken to predict the effects of Z. multiflora essential oil (EO), temperature and pH on the probability percentage of growth initiation of Staphylococcus aureus using artificial neural network. For validating the artificial neural network, we used test data which were not utilized for training. For determining the ability of artificial neural network, we obtained the Pearson's correlation between artificial neural network output and experimental log P% (r = 0.998 and P < 0.0001). In conclusion, taking into consideration the importance of S. aureus in food microbiology and the antimicrobial effects of the EOs which are commonly used for flavoring, the development of artificial neural network models are beneficial in order to predict the effects of Z. multiflora EO, temperature and pH on the probability percentage of growth initiation of S. aureus. Staphylococcus aureus is an important pathogen in food safety. Staphylococcal food poisoning is widespread and quite frequent. It is also among the four most common causes of foodborne illnesses. One of the most important purposes of food safety is inhibition of this microorganism growth. Substitution of traditional food preservatives by natural ones is a growing interest in food safety. Essential oils are aromatic oily liquid obtained from plant material. Plant essential oils (EOs) have varying degrees of antimicrobial activity. On the other hand, predictive microbiology is an essential element of modern food microbiology and offers to provide a scientific foundation to meet the ongoing needs of food safety. Over the last few years, artificial neural networks have been proposed as nonlinear modeling techniques in predictive microbiology.