Modelling of Milk Kefir Fermentation for its Optimised Physicochemical and Microbiological Properties

Milk kefir is a tangy, probiotic-rich fermented milk drink that is underexplored, partly due to the lack of broad awareness and complexity in standardisation of processing. This research studied the milk kefir fermentation process using response surface methodology (RSM) and artificial neural network (ANN). The RSM approach has optimised milk kefir quality at pH 4.4, viscosity at 1200 mPa.s and lactic acid at 0.8% at fermentation temperature of 35.8°C for 8.8 hours. The ANN model presented predictive capabilities with a higher coefficient of determination, R2 values and comparably lower root mean square error (RSME) and lower average absolute deviation (AAD) as compared to RSM, suggesting that the ANN model is more effective in capturing nonlinear data for predicting quality responses of milk kefir. Bacterial and fungal composition of milk kefir was measured using metagenomics via next-generation sequencing. Firmicutes was found as the most dominant bacterial phylum, while Ascomycota, the fungi phylum in both optimised and commercial milk kefir. At the genus level, Lactobacillus was an abundant bacterium in optimised milk kefir, while Streptococcus in commercial milk kefir. Both Lactobacillus and Streptococcus are recognised as probiotics that promote improvement in gut health and support immunity. For fungi, the genus Pichia was detected in high abundance percentage in optimised milk kefir, while Debaryomyces in commercial milk kefir.

Text 125027.pdf - Published Version Download (574kB) Official URL or Download Paper: http://www.pertanika.upm.edu.my/resources/files/Pe...

Actions (login required)

View Item