Results
The results ofcircRNA identification and annotation
The reads were aligned to the reference genome using bwa mem software. The read distribution of OS and corresponding adjacent nontumor tissues on the chromosomes were shown in Supplementary Figure 1a and 1b, separately. After bwa mem alignment, sam files were subjected to CIRI2 processing twice. Firstly, the junction reads were detected by paired chiastic clipping (PCC) signals. Basing on paired-end mapping (PEM) and GT-AG sequence feature, the reads were preliminary filtered to obtain candidate circRNAs. Secondly, the junction reads were detected again to filter out the false-positive candidate circRNAs (Supplementary Figure 1c). Totally, 6646 circRNAs were identified in OS and nontumor tissues, and in each type of samples, with 1-3 circRNAs on most genes. Among these 6646 circRNAs, 6014 circRNAs were distributed on exon region, 459 circRNAs were distributed on intron region, and the rest 173 circRNAs were distributed on intergenic region. Moreover, 4275 circRNAs have been annotated before, and the remaining 2371 were newly annotated. The lengths of 6646 circRNAs were mainly distributed in the range of 150-1000 bp (Supplementary Figure 1d).