CIRCexplorer2 assemble carries out de novo assembly for circular RNA based on RABT method.

Usage and option summary


CIRCexplorer2 assemble [options] -r REF <circ_dir>


-h --help                      Show help message.
-v --version                   Show version.
-r REF --ref=REF               Gene annotation file.
-p THREAD --thread=THREAD      Running threads. [default: 10]
--bb                           Convert assembly results to BigBed.
--tophat-dir=TOPHAT_DIR        TopHat mapping directory.
--chrom-size=CHROM_SIZE        Chrom size file for converting to BigBed.
--remove-rRNA                  Ignore rRNA during assembling (only for human hg19).
--max-bundle-frags=FRAGMENTS   Cufflinks --max-bundle-frags option.

Notes about options

  1. CIRCexplorer2 assemble require to use alignment results of TopHat2 to de novo assemble circular RNA transcripts. So if the TopHat2 result folder is not under <circ_dir>, you could specify its path using --tophat-dir.
  2. CIRCexplorer2 assemble will search for the chrom size file in TopHat2 result folder, and you could also specify its path using --chrom-size. If you have used CIRCexplorer2 align to align RNA-seq data, the chrom size file would be existed in the TopHat2 result folder.
  3. Assembly for rRNA would be very time-consuming. If you set --remove-rRNA option, it would skip assembly for rRNA. To be noted, this option is only suitable for hg19. If the assembly step is still very slow, you could set --max-bundle-frags with a small number. Please see Cufflinks protocol for more details about --max-bundle-frags option.
  4. If you set --bb option, the BigBed file of assembled transcripts would be created.


CIRCexplorer2 assemble needs a gene annotation file and a <circ_dir> folder which contains fusion_junction.bed created by CIRCexplorer2 parse or CIRCexplorer2 align. If there is no tophat folder under <circ_dir>, you should indicate its path via --tophat-dir. The gene annotation file should be in the format of Gene Predictions and RefSeq Genes with Gene Names. See Annotate for more details.


CIRCexplorer2 assemble will create one cufflinks folder under the <circ_dir> folder. It will also create cufflinks.log under <circ_dir> folder. The transcripts_ref.txt would be used to do alternative splicing analysis for circular RNAs, and it has the same format with refFlat format.

├── filtered_junction.gtf
├── filtered_junction.txt
├── genes.fpkm_tracking
├── isoforms.fpkm_tracking
├── skipped.gtf
├── transcripts.gtf
├── transcripts.txt
├── transcripts_ref.txt
├── transcripts_ref.bed
├── transcripts_ref_sorted.bed
  • filtered_junction.gtf and filtered_junction.txt: Filtered gene annotation files.
  • genes.fpkm_tracking, isoforms.fpkm_tracking, skipped.gtf and transcripts.gtf: Cufflinks result files.
  • transcripts_ref.txt, transcripts_ref.bed and transcripts_ref_sorted.bed: Circular RNA transcript files.
  • BigBed file of Circular RNA transcripts.