Analysis Of Sourdough Microbial Diversity Based On High-Throughput Sequencing And Machine Learning

Sourdough is a traditional fermented product made with grain processing products and water as the main raw materials,which is formed by natural fermentation or cultivation and fermentation after inoculation with microorganisms.In this study,sourdough from Henan,Shandong,Shanxi and Shaanxi province was collected as the research object,culturedependent method and high-throughput sequencing technology were used to analyze the microbial community structure of sourdough,clustering and dimensionality reduction algorithms in machine learning were used to analyze the correlation between moisture,geographical location and microbial community structure of sourdough,random forest algorithm was also used to identify key microorganisms,in order to provide theoretical reference for the industrial production of sourdough.Firstly,17 species of lactic acid bacteria(275 strains in total)were isolated from the65 sourdough by culture-dependent method,including Fructilactobacillus sanfranciscensis,Bacillus amyloliquefaciens,Lactiplantibacillus plantarum,Leuconostoc citreum,Enterococcus faecium,Weissella cibaria,Pediococcus pentosaceus,Companilactobacillus crustorum,Weissella confusa,Leuconostoc mesenteroides,Levilactobacillus brevis,Latilactobacillus curvatus,Lacticaseibacillus paracasei,Pediococcus acidilactici,Latilactobacillus sakei,Furfurilactobacillus rossiae and Leuconostoc fallax.Among them,Fructilactobacillus sanfranciscensis and Lactiplantibacillus plantarum were the dominant species,accounting for 18.2% and 15.3%,respectively.9 species of yeast(160 strains in total)were isolated,including Saccharomyces cerevisiae,Wickerhamomyces anomalus,Pichia membranifaciens,Hanseniaspora opuntiae,Kazachstania humilis,Meyerozyma guilliermondii,Torulaspora delbrueckii,Saccharomycopsis fibuligera and Pichia kudriavzevii.Among them,Saccharomyces cerevisiae and Wickerhamomyces anomalus were the dominant yeast,accounting for 37.5% and 23.1%,respectively.Secondly,the results of high-throughput sequencing technology showed that in sourdough,the mainly fungi were Saccharomyces cerevisiae,Kazachstania humilis,Saccharomycopsis fibuligera and Wickerhamomyces anomalus,80% of the sourdough had the Saccharomyces cerevisiae,which was the dominant species.There were obvious differences in the composition of fungal communities in sourdough from different regions,The relative abundance of Kazachstania humilis in sourdough in Shaanxi province was the highest,while the Saccharomycopsis fibuligera in Shanxi was the highest.The mainly bacteria in sourdough were mainly Fructilactobacillus sanfranciscensis,Pediococcus pentosaceus,Weissella cibaria,Lactiplantibacillus plantarum and Leuconostoc citreum,Fructilactobacillus sanfranciscensis as the dominant species.Compared with other regions,sourdough from Henan province has more Pediococcus pentosaceus,while Weissella cibaria was relatively abundant in Shaanxi province.Finally,based on the machine learning algorithm,the correlation between moisture content,geographical location and the microbial community of sourdough was explored,as well as the identification of key microorganisms in sourdough.It was found that moisture content in sourdough had a greater influence on microbial diversity,while geographical factors had less influence.The key lactic acid bacteria in the sourdough with moisture content above 15% were Fructilactobacillus sanfranciscensis and Leuconostoc citreum,Pediococcus pentosaceus and Levilactobacillus brevis were mainly in the sourdough with moisture content below 15%.The identification results of key microorganisms in sourdough from different regions showed that the key lactic acid bacteria in Henan province were Leuconostoc citreum,Levilactobacillus brevis and Latilactobacillus curvatus,Saccharomyces cerevisiae and Issatchenkia orientalis were the key yeast.Saccharomycopsis fibuligera were the key fungi in Shandong sourdough.The key lactic acid bacteria in Shanxi sourdough were Pediococcus pentosaceus and Lactobacillus paralimentarius.Fructilactobacillus sanfranciscensis and Kazachstania humilis were the key lactic acid bacteria and yeast in Shaanxi sourdough.