ecc_caller

ecc_caller is a pipeline designed to take Illumina NovaSeq paired end sequencing reads generated from isolated circular DNA and call eccDNA forming regions.

The pipeline uses information from two structural read variants to call eccDNA forming regions: split reads and opposite facing read pairs.

It also assigns a confidence score based off the number of split reads corresponding to each region as well as its coverage.

More detail on the pipeline is included in this manuscript: The extrachromosomal circular DNAs of the rice blast pathogen Magnaporthe oryzae contain a wide variety of LTR retrotransposons, genes, and effectors

Software requirements

ecc_caller was written to run in a RedHat Enterprise Linux environment with GNU bash version 4.2.46(20)-release.

ecc_caller was written using the following software and versions. It is likely that ecc_caller will work with newer versions of these software but it is not guaranteed. Make sure these are added to PATH:

• cutadapt version 2.4
• BWA MEM version 0.7.17-r1188
• samtools version 1.8
• bedtools version 2.28.0
• GNU parallel version 20180322
• Picard tools version 2.9.0 (with java version 1.8)

ecc_caller contains Python scripts which were written in Python 3.7.4 and requires the following modules:

• pandas version 0.25.1
• numpy version 1.17.2
• scipy version 1.4.1

Once all software is installed with paths set and the git repository has been cloned make sure to set this path variable so the wrapper scripts know where to find the python scripts and your picard.jar

export ECC_CALLER_PYTHON_SCRIPTS=/path/to/git/repo/python_scripts/
export ECC_CALLER_PICARD=/path/to/picard/

Conda installation instructions

Here are some example instructions for installation through conda

# create conda environment and activate
conda create -n ecc_caller
conda activate ecc_caller

# install required software
conda install python=3.7.4 pandas=0.25.1 numpy=1.17.2 scipy=1.4.1

conda install -c bioconda cutadapt=2.4 \
    bedtools=2.28.0 bwa=0.7.17 samtools=1.7

conda install -c conda-forge parallel=20180322

Then, download and unzip the picard files from here: https://github.com/broadinstitute/picard/releases/tag/2.9.0, and follow the instructions in the README for how to set up picard tools.

Next you'll want to create a directory to install ecc_caller and go into it. Then, to clone this git repository simply use:

git clone https://github.com/pierrj/ecc_caller.git

Finally, set paths, including for ecc_caller:

export ECC_CALLER_PYTHON_SCRIPTS=/path/to/install/directory/ecc_caller/python_scripts/
export ECC_CALLER_PICARD=/path/to/picard/
export PATH="/path/to/install/directory/ecc_caller/:$PATH"

You will need to define these environmental variables every time you run ecc_caller. Or you could set them every time you open a shell by adding these lines to your bash profile like so:

echo "export ECC_CALLER_PYTHON_SCRIPTS=/path/to/git/repo/python_scripts/" >> ~/.bash_profile
echo "export ECC_CALLER_PICARD=/path/to/picard/" >> ~/.bash_profile
echo "export PATH=/path/to/install/directory/ecc_caller/:$PATH" >> ~/.bash_profile

Usage

ecc_caller uses a mapfile to work with any input genome. This mapfile should be a list of the first field of all fasta entries (scaffolds/chromosomes) of interest for the analysis. This can be all scaffolds in the original genome file or only scaffolds of interest (i.e. excluding mitochondria).

Example command to create mapfile:

grep '>' guy11_genome_baoetal2017_with_70-15_mito.fasta | grep -v mito | awk '{print substr($1,2)}' > mapfile

Before mapping, remember to create a bwa index of your genome:

bwa index -p genome_bwa genome.fasta 

Then remove adapters and map sequencing reads. NOTE: "output_name" should be a string here, not a path to file, otherwise this command will fail.

generate_bam_file.sh -g genome_bwa -1 R1.fastq -2 R2.fastq -s output_name -t n_threads -m mapfile

Next call eccDNA forming regions

call_ecc_regions.sh -m mapfile -s output_name -t n_threads -b uniq.filtered.sorted.output_name.bam -q multimapped.filtered.name_sorted.output_name.bam

Finally, assign confidence to each eccDNA forming region

assign_confidence.sh -m mapfile -s output_name -t n_threads -b no_secondary.filtered.sorted.output_name.bam -r output_name.confirmedsplitreads.bed

All output files are written to the current directory.

Outputs

ecc_caller currently outputs many useful files for analysis and QC. File names need to be cleaned up a bit, update coming soon.

A few important files to note:

• filtered.sorted.output_name.bam - output from bwa mem, sorted, and filtered to scaffolds of interest
• lengthfiltered.merged.splitreads.output_name.renamed.bed - bed file containing split reads to be used for ecc calling
• outwardfacing.output_name.bed - bed file containing all outward facing read pairs to be used for ecc calling
• outputname.confirmedsplitreads.bed - output from call_ecc_regions.sh, split reads confirmed by the presence of an outward facing read pair
• output_name.ecc_caller_out.details.txt - tsv containing location of eccDNA forming regions as well as their confidence score, reason for confidence score and depth of coverage (see below). Columns are, in order: chromosome, start coordinate, end coordinate, number of split reads, confidence score, explanation of confidence score call, average number of reads in region of the length of the eccDNA before; overlapping; and after.
• output_name.ecc_caller_out.genomebrowser.bed - bed file of all eccDNA forming regions, colored by confidence score (see below)
• output_name.ecc_caller_out.splitreads.bed - bed file with one entry per split read supporting an eccDNA forming region with a confidence score of conf or hconf

Confidence Scores for eccDNA Forming Regions

coverage_confirm_nodb.py assigns confidence scores to each eccDNA region. These criteria can currently only be modified by changing the python script but I will eventually include this as a command line argument. These are as follows:

• lowq - low quality, greater than 5% of the region is uncovered by mapped reads and/or only one split read is associated with this region (colored red in bed file output)
• conf - medium confidence, 2 split reads associated with this region, coverage of the region is less than double that of the flanking regions of equal sizes (colored yellow in bed file output)
• hconf - high confidence, 2 split reads and coverage of the region is more than double that of the flanking regions of equal sizes OR 3 or more split reads associated with this region (colored green in bed file output)

Acknowledgements

Initial framework for the pipeline was inspired by the methods described in the following paper but all code in this repository is original:

Møller, H.D., Mohiyuddin, M., Prada-Luengo, I. et al. Circular DNA elements of chromosomal origin are common in healthy human somatic tissue. Nat Commun 9, 1069 (2018). https://doi.org/10.1038/s41467-018-03369-8

Thank you to Ksenia Krasileva and all members of the KVK lab for their feedback on progress on this pipeline.

License

ecc_caller is freely available under the MIT license

Contact

Please feel free to contact me directly with any questions

Pierre Joubert - @pmjoubert - pierrj@berkeley.edu

Project Link: https://github.com/pierrj/ecc_caller