Rapid screening and identification strategy of lactic acid bacteria and yeasts based on Ramanomes technology and its application in fermented food

There is an urgent need for rapid, sensitive, comprehensive and low-cost quality assessment strategies for probiotic resource development. we prepared AgNPs nanostructure array (AgNPs NSA) chips with stable signal and good enhancement effect as the substrate material, and combined with SERS technology to establish a Ramanomes reference of five species of lactic acid bacteria and three species of yeasts. Furthermore, we combined the omics research tool MetaboAnalyst 5.0 to analyze the biological information and the main difference components of Raman spectrum characteristics, and determined the qualitative identification ability and detection sensitivity by using the intuitive and efficient systematic cluster analysis (SCA) and convolutive neural networks (CNN) machine learning algorithm. Ultimately, we constructed a rapid lactic acid bacteria and yeasts isolation and enrichment method applicable to different kinds of food media without cultivation, and validated its recovery rate and the accuracy of qualitative modeling for SERS detection. In this study, the rapid isolation, purification and identification of probiotics in complex food matrices were investigated systematically, in order to break through the limitations of the current traditional model of “cultivating before screening” and provide a good technical support for the mining of probiotic resources. [Display omitted] •Rapid identification of probiotics by combining Ramanomes and SERS technology.•Establish a preliminary screening strategy of “screening before cultivating” probiotics.•Chemometric analysis differentiated chemotypes of lactic acid bacteria and yeasts.•The accuracy of CNN recognition was 100 %, and intercellular heterogeneity was quantified. There is an urgent need for quick, sensitive, thorough, and low-cost preliminary screening and rapid identification method for probiotic resource development. Here, we constructed a culture-free, accurate and sensitive “separation-enrichment-detection” rapid evaluation and identification method of probiotics in fermented food based on Ramanomes technology, and established a Ramanome reference of lactic acid bacteria and yeasts by AgNPs nanostructure array (AgNPs NSA) chips with uniform “hot spots” and high signal enhancement ability as SERS substrates. Systematic cluster analysis could clearly distinguish among different genera of lactic acid bacteria and yeast, and could indicate the affinity and difference between lactic acid bacteria of the sam