The Role of Dysbiosis of Gut Microbiota in the Pathogenesis of PCOS.

Data quality assurance: ① all inspections and measurements will be performed by either the hospital or the sequencing company personnel according to standard operating procedures (SOPs), except for saliva and stool samples, which will be self-collected by patients. For sample collection, we will provide text descriptions of the SOPs as well as video instruction. Designated staff will be assigned for support and can be contacted if participants have any queries concerning sample collection; ② a case report form (CRF) will be prepared according to the current SOPs, and detailed instructions will be provided to ensure consistency in data collection. At the same time, each CRF will be properly stored at least 5 years for verification and backtracking; ③ all experimental data will be logged into the database to ensure information accuracy based on the existing data; ④ we will keep the contact information of each participant, remind them of precautions during participation, and conduct regular follow-ups.

Sample size determination: The number of participants is based on comparable sample sizes in the literature. In this trial, there will be 50 healthy individuals (control group) and 150 PCOS (polycystic ovary syndrome) patients. The 150 PCOS patients will be randomly assigned to three intervention groups. This sample size accounts for a plausible insufficiency of data caused by patient dropouts and withdrawals before the study is completed. The participation cycle is of approximately four months, followed by a 2-year follow-up.

Metagenomic sequencing technology Metagenomic sequencing is the main technique used in this study. Metagenomics, also known as economics, was first proposed by Handelman and studies the molecular composition of microbial populations, their interactions, and gene functions.

In medicine, metagenomics compares the structural and functional changes of human microbial communities under normal and disease states. It can analyze the diversity and the functional differences of microbial communities from healthy individuals and from patients with diseases, thus determine how diseases relate to changes in the microbial communities and in their respective metabolic networks. Therefore, metagenomics provides theoretical evidence for disease prevention, detection, and treatment. At present, the internationally renowned Human Microbiome Project (HMP, http://www.hmpdacc.org/) and the Metagenomics of the Human Intestinal Tract (MetaHIT) are typical applications of metagenomics in medicine.

[Metagenomic species, genes, and functional annotation]

① Data quality control: the sequenced raw data will contain a certain amount of low-quality data, so quality control must be performed. Only high-quality data can correctly reflect the actual occurrence of microorganisms in the sample.

② Metagenome assembly: based on Clean Data, individual samples will be assembled separately at first, then reads that do not participate in the assembling above will be combined and mixed for assembly. This will increase the sequencing depth of low-abundance species in each sample and provide more sequencing information for each species.

③ Gene prediction: MetaGeneMark will be used for gene prediction based on single samples and mixed-assembled scaftigs. The redundancy of all predicted genes will be reduced to obtain a Uniq gene set. Then, the Clean Data of each sample will be compared to the gene set and the abundance of the gene set will be determined for each sample.

④ Species annotation: Clean Data will be used for quality control. It will be compared with an annotated according to reference genome databases of bacteria, archaea, viruses, and fungi from NCBI. A species abundance table will be obtained for each sample at different classification levels.

⑤ Functional annotation: functional annotation and abundance statistics will be based on the Uniq gene set and the KEGG database.