ABSTRACT: The gastrointestinal tract, the largest human microbial reservoir, is highly dynamic. The gut microbes play essential roles in causing colorectal diseases. In the present study, we explored potential keystone taxa during the development of colorectal diseases in central China. Fecal samples of some patients were collected and were allocated to the adenoma (Group A), colorectal cancer (Group C), and hemorrhoid (Group H) groups. The 16S rRNA amplicon and shallow metagenomic sequencing (SMS) strategies were used to recover the gut microbiota. Microbial diversities obtained from 16S rRNA amplicon and SMS data were similar. Group C had the highest diversity, although no significant difference in diversity was observed among the groups. The most dominant phyla in the gut microbiota of patients with colorectal diseases were Bacteroidetes, Firmicutes, and Proteobacteria, accounting for >95% of microbes in the samples. The most abundant genera in the samples were Bacteroides, Prevotella, and Escherichia/Shigella, and further species-level and network analyses identified certain potential keystone taxa in each group. Some of the dominant species, such as Prevotella copri, Bacteroides dorei, and Bacteroides vulgatus, could be responsible for causing colorectal diseases. The SMS data recovered diverse antibiotic resistance genes of tetracycline, macrolide, and beta-lactam, which could be a result of antibiotic overuse. This study explored the gut microbiota of patients with three different types of colorectal diseases, and the microbial diversity results obtained from 16S rRNA amplicon sequencing and SMS data were found to be similar. However, the findings of this study are based on a limited sample size, which warrants further large-scale studies. The recovery of gut microbiota profiles in patients with colorectal diseases could be beneficial for future diagnosis and treatment with modulation of the gut microbiota. Moreover, SMS data can provide accurate species- and gene-level information, and it is economical. It can therefore be widely applied in future clinical metagenomic studies.
