Primary analysis pipeline for ultra-accurate sequencing data

bwa v0.7.17 bedtools v2.26.0 samtools v1.7 snakemake v5.7.1 optional: conda 4.7.12

scikit-bio v0.5.5 biopython v1.74 optional: matplotlib v3.1.0 seaborn v0.9.0

Copy the git directory:

git clone github.com/jmcbroome/circleseq

Ensure external software dependences are installed and on your shell's path.

Package dependencies can be installed independently or the circleseq.yml environment may be used via conda.

conda create -c conda-forge -c bioconda -n circleseq snakemake scikit-bio biopython seaborn samtools bedtools bwa
conda activate circleseq

The snakefile as it stands expects input files in the format of {sample}_R1.fa and {sample}_R2.fa under the "input" file folder. Reference data is expected under references/{reference_genome}.fa, replacing bracketed values with the specific values of your sample and the reference genome name.

The file structure should look like this:

Directory with scripts
    input
         {sample}_R1.fq.gz
         {sample}_R2.fq.gz
    references
         {reference_genome}.fa

First, ensure your reference of choice is indexed with bwa.

bwa index references/{reference_genome}

Then simply call:

snakemake -j {max_threads} {sample}_{reference_genome}.txt

To include the final optional graphing step (dependencies are matplotlib and seaborn), instead call:

snakemake -j {max_threads} {sample}_{reference_genome}.png

Or run the graph_mutations.py script separately on the error table resulting from the above pipeline.

Again, replacing bracketed values with the name of your sample, the name of your reference genome file, and with the maximum threads value being the maximum number of threads available to the pipeline for processing. Default value for max_threads is 1. Add the argument "--use-conda circleseq.yml" as an alternative to global installation of requisite packages, or activate the environment with conda and call snakemake from within it.

Note that the mutations here may include unfiltered mismapping errors and similar. Downstream analysis should be generally performed using the constructed consensus bam, or the accompanying pileup variants.txt.

Reference and simulated input data have been provided to run an example to ensure correctly installed dependencies.

To call the test case, simply input at the command line:

snakemake -c1 simulated_yeast.txt