High-throughput RNA-seq
RNA library preparation was conducted by a kit of Truseq® chain total RNA sample preparation (Illumina, USA) based on provided directions, with the final library being evaluated with a Qubit 2.0 fluorometer and an Agilent 2100 Bioanalyzer for quantification and quality control. CBOT was then used to dilute the library to 10 pm for cluster generation, after which an Illumina Hiseq 2500 instrument (Illumina) was used for sequencing. All library construction and sequencing were performed by OG-Biotech Inc. (Shanghai, China).
For data analysis, RNA-seq read data were first subjected to quality control (QC) using FastQC (v. 0.11.3) (http://www.bioinformatics.babraham.ac.uk/projects/fastqc ), after which low-quality reads, rRNA sequences, and adapter sequences were trimmed with the seqtk tool (https://github.com/lh3/seqtk ). Mapping to the hg38 human reference genome was then performed with the BWA-MEM computer program (v 2.0.4), circRNAs were estimated by utilizing CIRI (33), while SRPBM was used to determine circRNA counts(34). Screening for circRNAs that were differentially expressed in AS and control samples was conducted utilizing edgeR using the following criteria: FC ≥ 2 or < 0.5 and P -value < 0.05. Functional characterization of circRNAs of interest was performed through GO and KEGG enrichment analyses of the parental genes from which these circRNAs were derived. The miRNA targets of these DECs were predicted with the miRanda software-based custom-built software application used by OG-Biotech(35), and Cytoscape was used for circRNA-miRNA network visualization(36).