CircMiner: Accurate and Rapid Detection of Circular RNA through Splice-Aware Pseudo-Alignment Scheme

A sensitive and fast computational tool for detecting circular RNAs (circRNAs) from RNA-Seq data.

Table of contents

1. Installation
2. Commands Options
3. Test Run
4. Example Commands
5. Output Files
6. Output Format
7. Contact & Support

Installation

BIOCONDA

CircMiner can be installed using conda package manager via bioconda channel:

conda install -c bioconda circminer

Installation from Source

Prerequisite. You will need (i) GCC 4.9.4 or higher; and (ii) GNU sort to compile the source code.

To install, you need to first fetch the git repository or download the corresponding compressed files.

git clone --recursive https://github.com/vpc-ccg/circminer.git
cd circminer
make 

Now you are ready to go!

Commands Options

Synopsis

circminer --index -r FASTA_FILE [OPTIONS]
circminer -r FASTA_FILE -g GTF_FILE -1 FASTQ_FILE_R1 -2 FASTQ_FILE_R2 [OPTIONS]

OPTIONS

Run circminer -h to see available options.

Indexing options

-i, --index:            Indicates the indexing stage.
-m, --compact-index:    Use this option only while building the index to enable compact version of the index.
-k, --kmer:        Kmer size [14..22] (default = 20).

General options:

-r, --reference:    Reference file.
-g, --gtf:        Gene model file.
-s, --seq:        Single-end sequence file.
-1, --seq1:        1st paired-end sequence file.
-2, --seq2:        2nd paired-end sequence file.

Advanced options:

-l, --rlen:        Max read length (default = 300).
-e, --max-ed:        Max allowed edit distance on each mate (default = 4).
-c, --max-sc:        Max allowed soft clipping on each mate (default = 7).
-w, --band:        Band width for banded alignment (default = 3).
-S, --seed-lim:        Skip seeds that have more than INT occurrences (default = 500).
-T, --max-tlen:        Maximum template length of concordant mapping. Paired-end mode only (default = 500).
-I, --max-intron:    Maximum length of an intron (default = 2000000).
-C, --max-chain-list:    Maximum number of chained candidates to be processed (default = 30).
-o, --output:        Prefix of output files (default = output).
-t, --thread:        Number of threads (default = 1).
-A, --sam:        Enables SAM output for aligned reads. Cannot be set along with --pam.
-P, --pam:        Enables custom pam output for aligned reads. Cannot be set along with --sam.
-d, --verbosity:    Verbosity mode: 0 or 1. Higher values output more information (default = 0).
-a, --scan-lev:        Transcriptome/Genome scan level: 0 to 2. (default = 0)
            0: Report the first mapping.
            1: Continue processing the read unless it is perfectly mapped to cDNA.
            2: Report the best mapping.

Other options:

-h, --help:    Shows help message.
-v, --version:    Current version.

Test Run

To make sure CircMiner is successfully installed and runable on your machine, you can download a small sample test here. Unzip the downloaded package and use the following commands to run CircMiner for building the index and calling circRNAs.

Example Commands

Indexing reference genome:

$ ./circminer --index -r ref.fa -k 20 --thread 4

Mapping to reference genome and circRNA calling:

$ ./circminer -r ref.fa -g ref.gtf -1 R1.fq -2 R2.fq -o output

After a successful run output.circ_report should contain the following two lines:

1	586821	608056	8	STC	CT-AG	CT-AG	Pass	Circ1-12,Circ1-56,Circ1-18,Circ1-50,Circ1-110,Circ1-80,Circ1-2,Circ1-4
1	805799	810170	9	STC	TT-TA	TT-TA	Pass	Circ2-32,Circ2-40,Circ2-36,Circ2-74,Circ2-76,Circ2-80,Circ2-60,Circ2-44,Circ2-2

Annotate CircMiner’s output file

To annotate output.circ_report with the corresponding gene, transcript, and exon numbers please run the following script:

$ python scripts/annotate_transcript.py output.circ_report ref.gtf output.circ_report.annotated

Converting GTF file

If you use UCSC GTF files please use the following script to convert them to Ensembl GTF style and use the converted GTF when running CircMiner.

$ python scripts/convertGTF.py INPUT_GTF OUTPUT_GTF

Output Files

When a successful run finishes, the structure of output directory (e.g. outdir) will be as follows:

outdir                                         # output directory
├── output.circ_report                         # detected circRNA report
├── output.candidates.pam                      # back-splice juntion read mappings
└── output.mapping.pam/output.mapping.sam      # pseudo-alignment mapping results (only available when corresponding argument is passed)

Output Format

The information regarding the detected circRNAs is reported in output.circ_report file. It includes the exact breakpoint location, number of supporting back-splice junctions and their read names.

Column	Type	Description
1	string	Chromosome name
2	int	Start genomic position of circRNA
3	int	End genomic position of circRNA
4	int	Number of supporting back-splice junction reads
5	string	Type of circRNA
6	string	Consensus of splice signal on supporting back-splice junction reads
7	string	Splice signal on reference
8	string	Pass/Fail (based on matching splice signal to reference)
9	string	Supproting back-splice junction read names (comma-separated)

The back-splice juntion read mappings are stored in output.candidates.pam. If –pam/–sam is specified in the input arguments, the mapping results will be available in output.mapping.pam/output.mapping.sam file.

Note: If the scan level parameter (-a, --scan-lev) is set to 2 while running the tool, the mapping with the smallest error and soft-clip values is reported.

PAM mapping format:

Column	Type	Description
1	string	Read name
2	string	Chromosome name (R1)
3	int	Start genomic position (R1)
4	int	End genomic position (R1)
5	int	Number of aligned basepairs (R1)
6	int	Start of aligned position on read (R1)
7	int	End of aligned position on read (R1)
8	char	Relative strand: “+” or “-“ (R1)
9	int	Edit distance (R1)
10	string	Chromosome name (R2)
11	int	Start genomic position (R2)
12	int	End genomic position (R2)
13	int	Number of aligned basepairs (R2)
14	int	Start of aligned position on read (R2)
15	int	End of aligned position on read (R2)
16	char	Relative strand: “+” or “-“ (R2)
17	int	Edit distance (R2)
18	int	Insert length
19	int	Number of junctions happening between two mates
20	int	Transcriptomic mapping: 1. Genomic mapping: 0

Contact & Support

For any bug report, feature request, or questions please fill out an issue through CircMiner’s issue page.

Citation

If you use CircMiner please cite our paper:

Asghari H., Lin YY., Xu Y., Haghshenas E., Collins CC., Hach F. CircMiner: Accurate and Rapid Detection of Circular RNA through Splice-Aware Pseudo-Alignment Scheme. Bioinformatics (2020). btaa232

Simulation Data

The simulated RNA-Seq reads used in this project can be accessed here.

Copyright and License

This software is released under GNU General Public License (v3.0).