Bioinformatic Pipeline for Processing lacI Library

Pipeline Overview

• This pipeline performs initial QC of the input PacBio sequencing data.
• The required input is a g-zipped fastq file contining PacBio CCS reads.
• Uses the lacI sequence to determine read direction and reverse complements reverse reads.
• Extracts target sequences, e.g. lacI and barcodes from reads.
• Generates QC stats for extracted reads.

Pipeline Environment

• Snakemake is used to compose the bioinformatic pipeline, https://snakemake.readthedocs.io/en/stable/.
• Snakemake and pipeline dependencies can be install using conda (https://www.anaconda.com/distribution/) and bioconda (https://bioconda.github.io/).
• The pipeline is defined in the Snakefile and adapter sequences are defined in the targets.csv config file.

Environment Setup

• Install conda with python 3.7 if not already installed on the system, https://docs.anaconda.com/anaconda/install/, miniconda as well as the full anaconda install will work.
• Generate a conda environment for running the pipeline, conda env create -n lacI_ccs --file environment.yaml.

Running pipeline

• Activate conda environment conda activate lacI_ccs.
• Run pipeline using snakemake -j [#]. The -j parameter takes the number of threads or parallel jobs snakemake executes simultaneously. (I like to also include -p, which prints the command snakemake executes at each step.) The dry run snakemake argument, -n, can be used to see which pipeline steps will be run.

Modifying pipeline

• Pipeline input files and output directory are defined in the config file, 'config.yaml'.
• The adapter sequences are defined and can be updated in the targets.csv file.
• Snakemake works similar to make in that it only reruns parts of the analysis pipeline when an input file is changed or the output file is missing. After updating the targets.csv file you will want to remove any pipeline output files you want to replace.

Test data

• The file, '/data_0/raw/test.fastq.gz', contains a suitable test dataset. To run the pipeline with another dataset, change the output directory (outdir) and input file (infq) in the config.yaml file.