Gut microbiome has gained wide attention for its close relationship with human health and has become the key target of precision medicine. As the second genome in hosts, the dysbiotic gut microbiome may cause various diseases, including digestive diseases (e.g., diarrhea, inflammatory bowel disease, colorectal cancer, etc.), metabolic diseases (e.g., hyperlipidemia, obesity, diabetes, etc.), immune diseases (e.g., asthma, rheumatoid arthritis, polyneuritis, etc.) and even neurologic diseases (e.g., autism, Alzheimer's syndrome, Parkinson's syndrome, etc.). With the accumulation of gut microbiome data, researchers have constructed many gut microbial databases and bioinformatic tools, providing great help for the following-up research. However, these databases and bioinformatic tools were developed under different scenarios, and their application scopes also vary. In this thesis, I introduced the common bioinformatic tools related to the gut microbiome analysis, and then applied gut microbiome analysis in four projects to improve our understanding of the diseases.

With the rapid development of sequencing technology, next-generation sequencing has become the main strategy to investigate the gut microbiome, due to its stable quality, high through-output, and low-cost. Since the gut microbiome is dominated by bacteria, many studies choose the 16S rRNA gene, which is a unique genetic sequence in prokaryotes, to detect the gut bacterial compositions. For the data produced by 16S rRNA sequencing, we can use four different databases for the taxonomical annotation, including Greengenes, SLIVA, RDP, and RefSeq. Also, platforms with different algorithms can be adopted for the 16S rRNA data aligning and the standard microbial analysis, such as Mothur, USEARCH, and QIIME2. However, since 16S rRNA mainly focuses on Bacteria and Achaea, more and more studies use metagenomic sequencing to detect the whole microorganisms. Metagenomic data contains all the microorganisms in the gut, and we can use it to detect the functional distributions of the gut microbiome. To promote the taxonomical annotation speed of the gut microbiome, researchers have piled up several metagenomic databases, such as the human gut microbial gene catalogue and the culturable genome references from BGI institute, the metagenomic species database from EBI institute, and the OpenBiome Microbiome Library from the Broad institute. Correspondingly, we can choose different tools for the metagenomic annotation, such as MetaPhlan3, Kraken2, and MEGAN6. Besides, we can establish the metagenomic genomic database by ourselves with metagenomic assembling tools, including MetaSpades, MEGAHIT, and OPERA-MS. These databases and tools provide a great convenience for gut microbiome research. Apply gut microbiome analysis for practical projects, we'd like to gain a deeper insight into gut microbiome in four diseases, including cerebral palsy and epilepsy, central precocious puberty, cholestasis, and undernutrition.

First, we explored the characteristics of gut microbiota in children with cerebral palsy and epilepsy. Gut microbiome affects the central nervous system with its metabolites, and associates with various neurologic diseases. However, the features of gut microbiome in patients with two compilated neurologic diseases remain unexplored. In this project, we adopted the USEARCH platform and RDP database to detected the gut microbial features in the children with cerebral palsy and epilepsy. For the patients, we discovered the enriched Enterococcus, Bifidobacterium, Clostridium IV, and Akkermansia. Since Akkermnsia increase the mucosal permeability by degrading the mucous layer, the dysbiotic gut microbiota would triggered systematic immunoreactions in hosts. These findings provided references for the bacterial adjuvant interventions for the neurologic diseases treatment, and further our understanding of the gut-brain axis.

Second, we compared the gut microbiota compositions between the healthy, over-weighted, and central precocious puberty girls. The 16S rRNA taxonomical annotation results suggested the altered bacterial compositions in the patients with central precocious puberty, such as Alistipes, Klebsiella, and Sutterella, and these bacteria have been reported in other neural diseases. Further detecting the functional changes of the gut microbiome, we discovered the enriched nitric oxide synthesis of the altered gut microbiome in both central precocious puberty and over-weighted participants. Since the elevated nitric oxide synthesis correlated with the levels of follicle-stimulating hormone and insulins, we provided clues for the occurrence of central precocious puberty from the perspective of gut microbiota.

Third, we investigated the alterations of gut microbiome in infants with cholestasis and detected the associations between the gut microbiota and hepatic functions. By comparing the gut microbiota between the healthy and cholestatis infants, we discovered the decreased Bifidobacterium, Bacteroides, and Faecalibacterium in the cholestatsis infants. Since these bacteria were associated with the enterohepatic circulation of bile acid, the dysbiotic gut microbiota might led to the disordered circulation, and caused cholestasis. Moreover, we constructed an infantile cholestasis-risk model with high accuracy, assisting the clinical prediction of cholestasis. Through this project, we have gained insight into the pathogenesis of infantile cholestasis, and provide a solid theoretical basis for gut microbial intervention on cholestasis treatment.

Fourth, we carried out gut microbiome exploration in undernutrition children, and established undernutrition-risk model based on the gut microbiome. As compared with the healthy children, the undernourished children exhibited altered gut microbial function, while one of the major features was the reduced iron transporters in the gut microbiome of the patients. The iron transporter assists the absorption of Fe2+ in hosts, and mainly came from Bacteroides uniformis. Therefore, these discoveries illuminated the pathogenies of undernutrition from the perspective of the gut microbiome. Based on the differentially enriched bacterial gene functions and gut bacteria, we obtained the undernutrition-risk models with area under curve (AUC) values of 0.87 and 0.91, respectively, assisting for the disease’s early prediction. 

In all, gut microbiome is a fascinating field under fast development, and it interacted with human health closely. The tools of gut microbiome facilitate us to gain insights into its roles in disease occurrence. Herein, we present the databases and tools for the 16S rRNA and metagenomic data analysis, and apply them to detect the pathogenesis of the diseases. We believe this thesis gives us a further step to understand neurologic, neuroendocrine, hepatobiliary, and metabolic diseases from the perspective of gut microbiome.