Analysis of Bacterial Diversity and Prediction of Phenotype in Rice Wine Koji from Different Regions

A rice wine koji collected from Nanning of Guangxi province and Xiaogan of Hubei province were used to sequence its 16S rRNA V₃~V₄ region based on Illumina Miseq high-throughput sequencing technology, combined with bioinformatics and multivariate statistical methods to analysis the bacterial diversity of rice wine koji in the two regions, while the bacterial taxa phenotypes were predicted. Results showed that, the Shannon index of the bacterial taxa in the rice wine koji from Nanning area was significantly lower than that of the Xiaogan area(P<0.05). Principal coordinate analysis based on weighted UniFrac distances showed a clear separation trend between the samples from the 2 regions in spatial arrangement. Multivariate analysis of variance indicated that the bacterial community structures of the rice wine koji in the 2 regions were significantly different(P<0.05). The bacterial taxa of rice wine koji in Nanning area were mainly Lactobacillus with an average relative abundance of 62.03%, while Weissella was predominant in the Xiaogan area with an average relative abundance of 50.14%. Phenotypic prediction showed that rice wine koji from Nanning area exhibited a significantly higher(P<0.01) abundance of gram-positive bacteria, while a significantly lower(P<0.01) abundance of oxidative stress tolerance, biofilm formation, gram negative, pathogenic potential and facultative anaerobic characteristics. It could be concluded that there were some differences in the bacterial taxa of rice wine koji from different regions.