DomainSeq

Author: Patrick V. Holec

Created: May 2018

Overview

This project is a pipeline for analysis of long-read enrichment processes. Specifically, if a library is generated composed of multiple domains, each member of each library in each selection cannot be observed easily through Illumina sequencing due to read-length restrictions. Additionally, PacBio provides the necesssary read-length, but is throughput limited. In these cases, multiple sources of sequencing data can be integrated to allow rapid, deep sequencing of libraries. Three types of files need to be provided:

1. An Excel document linking each possible domain to a DNA sequence
2. A PacBio .fastq file, containing long-reads connecting a particular barcode to the read of the DNA containing the domain set
3. An arbitrary number of Illumina sequencing files, containing short-reads which cover barcodes observed in specific stages during selection

DomainSeq is a pipeline that pieces these three sets of information together and presents processed visuals and documents describing the library state throughout selection. This takes the form:

Illumina data (BC) -> PacBio data (BC to DNA sequence) -> Excel document (DNA sequence to Domains)

A number of user controls are provided for each step of analysis along this analysis and can be scaled to any dataset size, taking appropriate steps to minimize computation time.

TODO:

• Multithreading

Requirements

This package has only been tested on Unix/Linux (OSX, Ubuntu, etc). The project is not platform specific, but installation instructions only apply to these systems. Additionally, all code is formated for Python 3.X.

Installation

We recommend using this package through a virtual environment. This creates an "instance" of Python that is isolated from the rest of your system. To do this, open terminal and do the following:

1. Clone repository by typing:

   git clone https://github.com/birnbaumlab/Kyung-et-al-2021.git

and move into the base directory (".../DomainSeq/") using change directory commands.

2. Install the virtual environment package to your native Python:

   python3 -m pip install --user virtualenv

3. Use this package to make a virtual environment of Python 3:

   python3 -m virtualenv domainseq-venv

   Note: This creates a folder in your current directory called "domainseq-venv" which contains your virtual environment. Call it whatever you want!

4. Activate your virtual environment by declaring it as your source for Python:

   source domainseq-venv/bin/activate

   Note: This command causes the virtual environment (your "domainseq-venv" folder) to be prioritized as the version of Python your session will use.

5. Install the required packages for the project:

   pip install -r requirements.txt

Great! If these commands worked without errors, you have:

• Copied the project
• Created a virtual Python environment
• Install the required packages for the project

If you ever want to update your current version, simply type:

git pull

which will update your local repository to the remote repository's latest version.

See the wiki for information on usage.