DBS-OCP

DBS for Open Chromatin Profiling in single cells/nuclei, a.k.a scATAC-seq/snATAC-seq.

• Install
• Running analysis
• Development

Install

1. Clone the DBS-OPC github

   git clone https://github.com/AfshinLab/DBS-OCP 
   

2. Install the base environment for environment.yaml file using conda:

   conda env create -n dbsocp -f environment.yaml 
   

3. Install the DBS-OPC package:

   pip install . 
   

To install on the Uppmax Bianca cluster follow these instructions.

Running analysis

Required inputs:

• Paired FASTQ files
• BWA indexed reference

1. Setup analysis directory called my_workdir:

   dbsocp init --read1 /path/to/read.1.fastq.gz my_workdir 
   

   This will create and populate the directory with symlinks to the paired FASTQ.
2. Update configs, here the path to the BWA indexed genome reference is added . To view current configs jsut run dbsocp config

   dbsocp configs -s reference /path/to/reference
   

3. To run the full analysis type the following.

   dbsocp run
   

   To only run the preprocessing steps (barcode correction, read trimming , barcode assignment and QC) include the argument preprocess.

Development

Information about ATAC-seq analysis can be found in the here.

Setup

To setup the development environment run:

conda env create -n dbsocp-dev -f environment.yaml

To install the package in editable mode with additional development dependancies run:

pip install -e .[dev]

Code style

Code style is enforced using black. To run the formatter:

black src/ tests/

Linting

Linting is done using flake8. To run the linter:

flake8 src/ tests/
flake8 --select=W292 --filename '*.yaml,*.yml'

Testing

Testing is done using pytest. To run the tests:

pytest -v tests