Population and ethics statement
This study was based on patients with otolaryngology and head and neck surgery in Zhongshan Hospital Affiliated to Xiamen University (February 2021 to February 2022). Institutional ethics committee approval for this project was provided prior to the start of this study and was in accordance with the Declaration of Helsinki. This study was performed with the approval of the Institutional Ethics Committee of Zhongshan Hospital Affiliated to Xiamen University (Approval number: 2021-077), and written and oral informed consents were obtained from patients or their relatives.
Subject hearing test
We performed hearing tests on randomly selected subjects. Three deaf patients and three healthy subjects were selected for inclusion in this study. All patients underwent audiometry (Otometrics, Model 1066) and audiograms were recorded. After the subjects were determined to be eligible for our study, the peripheral blood of the patients was drawn and temporarily stored in EDTA anticoagulant tubes.
PBMC isolation and rNA preparation
Patient blood was stored in EDTA anticoagulant tubes immediately after ex vivo and temporarily stored at 4℃. Take 3ml of blood from the EDTA anticoagulation tube into a 15ml freshly packaged centrifuge tube, add 3 times the volume of red blood cell lysate, place at 4℃ for 15min, centrifuge at 4℃ for 10min, draw the cell pellet after layering, add 1ml trizol to the cell pellet and mix well. Add 270ul chloroform to the centrifuge tube, vortex until the solution is emulsified and milky white, let stand for 5 minutes, centrifuge, suck the supernatant into a new EP tube, add an equal volume of isopropanol, mix well and let stand for 10 minutes, after centrifugation Remove the supernatant, add 50ul of nucleic acid-free plum water to the precipitation, and store the samples after fully dissolving.
MiRNA-seq sequencing was entrusted to Aimo Gene Xiamen Biotechnology Co., Ltd. MiRNA-seq experiments were performed on peripheral blood mononuclear cells provided by subjects. According to the manufacturer's instructions, the experimental protocol was performed according to standard procedures provided by Illumina, including library preparation and sequencing experiments. Small RNA sequencing libraries were prepared using TruSeq Small RNA Sample Prep Kits (Illumina, San Diego, USA). After the library preparation was completed, the constructed library was sequenced using Illumina novaseq6000, and the sequencing read length was single-ended 1X50bp. After the library is constructed, use Qubit2.0 for preliminary quantification, dilute the library to 1 ng/ul, and then use Agilent 2100 to detect the insert size of the library. After the insert size is as expected, use the Q-PCR method to determine the effective concentration of the library. Perform accurate quantification (library effective concentration > 2nM) to ensure library quality. The measured raw data (Raw Data) has a certain proportion of interference data (Dirty Data). In order to make the results of information analysis more accurate and reliable, data filtering is first performed on the raw data. Including removal of low-quality bases, linker sequences, and short sequences (less than 18bp). The raw data generated by sequencing needs to be preprocessed. We use TrimGalore to filter out unqualified sequences to obtain valid data (clean data). The reads obtained by sequencing were compared with the database, and the RNA types in the sequencing reads were identified. At the same time, in order not to affect downstream analysis as much as possible, we removed sequences identified as non-miRNAs. The subsequent differential expressed miRNAs (DEMs) were further analyzed.
MiRNA-seq data quality inspection was done with fastqc (0.11.7), data filtering was done with trim_galore (0.6.0), repeated sequence removal was done with (RepeatMasker) 4.0.9, and non-miRNA was removed with infernal (1.1.3), miRNA alignment was done using bowtie (1.0.0), predicted secondary domain was done using RNAfold (2.4.14), De novoe precursor scoring was done using miRDeep2 (220.127.116.11) , and differential expression was done using R:DEseq2/edgeR (1.24.0)  done. Statistical analysis, graphing was done using R (3.6.1) . Sequencing was performed using an Illumina novaseq6000. The original data processing process was completed by Xiamen Aimo Gene Biotechnology Co., Ltd. We further filtered hearing loss differentially expressed genes (HL-DEMs), the threshold were set at P-Value < 0.05 and |log FC|≥1, and the HL-DEMs obtained according to this criterion were included in our study. We selected the most up-regulated miRNAs for corresponding predictions. Aims to find downstream mRNAs.
MiRNA target gene prediction
The miRNA-mRNA was identified through the miRDB (http://www.mirdb.org/), mirDIP (http://ophid.utoronto.ca/mirDIP/index.jsp#r), mirtarbase (http://mirtarbase.cuhk.edu.cn/php/search.php#target) and TargetScan (http://www.targetscan.org/vert_80/) databases. Genes appearing in all four databases were considered target genes for DEMs. The miRDB database was used for miRNA target prediction and functional annotation. Transcriptome-wide target prediction data is available in the miRDB, including 3.5 million predicted targets regulated by 7000 miRNAs in 5 species . mirDIP v4.1 provides nearly 152 million human microRNA target predictions from 30 different sources . miRTarBase provides comprehensive information on experimentally validated miRNA-target interactions (MTIs) . The Targetscan database is used to predict target genes downstream of miRNAs .
Go and pathway enrichment analysis of target gene
Metascape  (http://metascape.org) is a premium database that introduces a workflow that integrates gene annotation, membership analysis, and meta-analysis of polygenic lists. It’s rich set of analysis tools is accessible through a convenient one-click quick analysis interface, and results are communicated through article-like analysis reports.
MCODE analysis, key candidate genes from PPI network
The STRING  (https://string-db.org/) database is designed to integrate all known and predicted associations between proteins, including physical interactions and functional associations. Among these associations, protein-protein interactions are particularly important due to their versatility, specificity, and adaptability. Minimum required interaction score: medium confidence (0.400). Network display options: hide disconnected nodes in the network.
Through the STRING database, we further screened the predicted downstream target genes, and finally obtained hub target genes. We predict that miRNAs may affect hearing loss by regulating hub target genes.