A software package for identifying eccDNAs in the ONT sequencing data of RCA-amplified eccDNA
A new bioinformatics pipeline called ECCFP has been developed to improve the detection of extrachromosomal circular DNA (eccDNA) from long-read sequencing data. ECCFP uses all consecutive full passes from individual reads for candidate eccDNA identification and consolidates candidate eccDNAs to generate accurate unique eccDNA. To thoroughly evaluate the performance of ECCFP, 6 simulated datasets, 2 real sequencing datasets with spiked-in artificial synthesized eccDNAs and 29 publicly available real sequencing datasets, were applied. The results demonstrate that ECCFP markedly reduces the false positive rate, significantly increases the number of unique eccDNAs detection, achieves higher accuracy in identifying junction positions and circular consensus sequences, and also shortens the runtime, when compared to other three eccDNA analysis pipelines.
 |
|---|
ECCFP is optimized from the Flec software.
git clone https://github.com/WSG-Lab/ECCFP.git
cd ECCFP
conda env create --file environment.yml -y
conda activate ECCFP
pip install -e .
# Verify installation and display help message
eccfp --help
eccfp --fastq input.fastq --paf mapping.paf --reference ref.fa -o output
usage: eccfp [-h] [-v] [--verbose] --fastq FASTQ --paf PAF --reference REFERENCE [--output OUTPUT] [--max-offset MAXOFFSET]
[--min-map-qual MINMAPQUAL] [--fluctuate FLUCTUATE] [--min-fullpass NF] [--min-coverage COV] [--min-candidates NC]
[--min-depth MINDP] [--min-allele-freq MINAF] [-t NUM_WORKERS] [--batch-size BATCH_SIZE]
ECCFP: EccDNA Caller based on Consecutive Full Pass
options:
-h, --help show this help message and exit
-v, --version show program's version number and exit
--verbose Show detailed debug logs
Input files:
--fastq FASTQ Input FASTQ format reads file
--paf PAF Input PAF format alignment file
--reference REFERENCE
Reference genome FASTA format file
--output OUTPUT, -o OUTPUT
Output directory, default is current directory (default: .)
PAF filtering:
--max-offset MAXOFFSET
Maximum offset between the start/end positions of two sub-reads (default: 20)
--min-map-qual MINMAPQUAL
Minimum mapping quality of a sub-read (default: 30)
Candidate eccDNA refinement:
--fluctuate FLUCTUATE
Maximum offset of candidate eccDNAs within the same group (default: 20)
--min-fullpass NF, -nf NF
Minimum number of fullpass reads supporting a candidate eccDNA (default: 2)
--min-coverage COV, -cov COV
Minimum fullpass count for merging candidate eccDNAs (default: 1)
--min-candidates NC, -nc NC
Minimum number of candidate eccDNAs (default: 2)
Consensus sequence and variant detection:
--min-depth MINDP, --minDP MINDP
Minimum depth for variant detection (default: 4)
--min-allele-freq MINAF, --minAF MINAF
Minimum allele frequency for variant detection (default: 0.75)
Performance control:
-t NUM_WORKERS, --threads NUM_WORKERS, --workers NUM_WORKERS
Number of parallel processes, 0 means auto-detect (cpu_count - 1, current machine: 31)
--batch-size BATCH_SIZE
Phase 1 batch size, 0 means auto-calculate (recommended for current machine: 14000) (default: 0)
Examples:
eccfp --fastq reads.fastq --paf mapping.paf --reference ref.fa -o output/
cd example
minimap2 -cx map-ont ref.fa example.fastq --secondary=no -t 8 -o mapping.paf
eccfp --fastq example.fastq --paf mapping.paf --reference ref.fa -o output
Five files are generated following the completion of the pipeline: unit.txt and candidate_consolidated.txt(intermediate files), final_eccDNA.txt, consensus_sequence.fasta, and variant.txt (result files).
file |
details |
|---|---|
| unit.txt | full pass alignment details for all candidate eccDNAs detected in reads |
| candidate_consolidated.txt | the consolidating steps used to derive accurate eccDNAs from the candidate eccDNAs |
| final_eccDNA.txt | accurate eccDNA information |
| consensus_sequence.fasta | the consensus sequences of accurate eccDNAs |
| variants.txt | variant profiles specific to these accurate eccDNAs |
| eccDNApos | Nfullpass | Nfragments | Nreads | refLength | seqLength |
|---|---|---|---|---|---|
| chr16:757924-758274(+) | 4 | 1 | 1 | 351 | 350 |
| chr1:21828265-21828439(+)|chr7:132167689-132167880(+) | 24 | 2 | 2 | 367 | 367 |
description |
|
|---|---|
| eccDNApos | eccDNA position |
| Nfullpass | Number of consecutive full pass for this eccDNA covered by all reads |
| Nfragments | Number of fragment that form this eccDNA |
| Nreads | Number of reads identified for the eccDNA |
| refLength | The length of reference genome that this eccDNA |
| seqLength | The length of consensus sequence that this eccDNA |
description |
|
|---|---|
| col1 | chromsome |
| col2 | position in the reference genome |
| col3 | reference base |
| col4 | variant |
| col5 | supportive coverage depth |
| col6 | total coverage depth |
| col7 | type |
| col8 | eccDNApos |
- Primary publication (iMetaOmics, 2026):
Li, W., Miao, B., Zhang, J., Zeng, Q., Zhang, T., Wu, Z., Song, Y., Li, M., Guo, L., Luo, J., Xu, J., Liu, T., Chen, S., Gu, J. and Wan, S. (2026), ECCFP: A consecutive full pass-based bioinformatic analysis for eccDNA identification from long-read sequencing data. iMetaOmics e70080. https://doi.org/10.1002/imo2.70080
- Protocol (Bio-protocol, 2026):
Li, W., Miao, B. and Wan, S. (2026). A Bioinformatics Workflow to Identify eccDNA Using ECCFP From Long-Read Nanopore Sequencing Data. Bio-protocol 16(6): e5636. DOI: 10.21769/BioProtoc.5636.