Jinyang Zhang

Hi Chee How, CIRI-full doesn't support trimmed reads with variable sequence length, please use raw reads instead. -Jinyang

Hi Dario, I think the overlapping reads in your libraries are causing the problem. These overlapping BSJ reads will have distinct alignment pattern than normal paired-end reads, and thus could be filtered out as false positive. Have you tried using reads merging tools (e.g. FLASH) to merge overlapping short reads, and run CIRI2 on merged and unmerged reads respectively, then use the sum of BSJ reads in two parts of reads as final results? I think this could be a convenient solution.

Hi Rabbani, Sorry for the inconvenience. Could you send the GTF file or first 100 lines of it to my email zhangjinyang@biols.ac.cn ? I will check why CIRI2 failed to load your annotation file. Thanks, Jinyang

Hi Rabbani, I am not sure what cause the problem. If the first few lines of your GTF file contain header information start with '#', you can remove them and try again. P.S. Are you using the latest version (v2.0.6) of CIRI2? It seems that the error message is generated from an obsolute version. Thanks, Jinyang

Hi Rabbani, Please check the attribute column (column 9) in your reference.gtf. CIRI2 need the "gene_id" attribute for circRNA annotation. If your annotation file do not provide this information, you need to run CIRI2 without it.

Hi Rabbani, Please the attribute column (column 9) in your reference.gtf. CIRI2 need the "gene_id" attribute for circRNA annotation. If your annotation file do not provide this information, you need to run CIRI2 without it.

Hi Mohammad, Please the attribute column (column 9) in your reference.gtf. CIRI2 need the "gene_id" attribute for circRNA annotation. If your annotation file do not provide this information, you need to run CIRI2 without it.