Species and strain identification of lactic acid bacteria using FTIR spectroscopy and artificial neural networks

Abstract

Lactic acid bacteria (LAB) are a heterogeneous group of micro-organisms with high relevance for fermented food and beverages as well as human and animal health. Identification of LAB is complicated by species richness, a high degree of heterogeneity between species and the presence of very closely related species at the same time. In the present study we developed identification systems based on FTIR spectra and artificial neural networks (ANNs) for species identification of 92 LAB species out of nine genera and strain identification of probiotic and environmental Enterococcus faecium strains used as additives for animal feed. The reference databases established in this work permit reliable identification of LAB with accuracies of 93.2% at species and 97.1% at strain level for Ec. faecium, demonstrating the excellent suitability of FTIR spectroscopy as a technique applicable for simultaneous species and strain identification.