Snakemake Workflow for 4C-ker

4C-Workflow is a snakemake implementation for the 4C-ker library. This generalized workflow aims to automate the 4C data analysis process allowing you to get to your results faster. To that end, the workflow supports multiple baits and comparisons by simply specifying the baits and comparisons within the config file.

This workflow depends heavily upon Miniconda, R.4Cker, and Snakemake so if you want know what they are checkout the links below. Otherwise just follow the setup process and you should be able to get 4C-Workflow up and running.

• Miniconda
• R.4Cker Github
• 4C-ker Paper
• Snakemake

Contents

• Setup
  • 4C-Workflow
  • Analysis
• Example
  • Download Raw Fastq Files
  • Setup config.yaml
  • Run
• Acknowledgements

Setup

• 4C-Workflow guides you through the setup process of getting the various parts of the 4C pipeline working.
• Analysis will show you how to describe your experiment to 4C-Workflow thus enabling it to perform an accurate analysis.

4C-Workflow

1. git clone the 4C-Workflow directory into your local directory
2. Install Miniconda
3. Use the following command conda config --add channels <specific_channel> to add the following channels
   • conda-forge, inso, bioconda, r
4. Once the appropriate channels are added use the terminal to create a conda environment while inside the 4C-Workflow folder
   • conda create --name 4C-Workflow --file environment.txt
5. Within terminal perform the following commands to install R.4Cker:
   • source activate 4C-Workflow
   • R
   • library(devtools)
   • install_github("rr1859/R.4Cker", ref = "b0e1923")

You are now ready to use 4C-Workflow to analyze your raw 4C data.

Analysis

1. Snakemake uses the config.yaml file to understand the experimental parameters, additional details in example config.yaml.

2. Lastly add both a .fasta reference genome and primary enzyme sequence to the reduced_genome folder

   • Make Sure the names for both the reference genome and enzyme match the names provided for primary_enz_name and reduced_genome from other
     • eg. dm6 == dm6.fasta & hindiii == hindiii.fasta

Note: everything you need to use 4C-Workflow is in the example config.yaml file, for additional help go through the example.

Example

The example config.yaml is setup to handle an experiment with two baits, someBait1 and someBait2, and three comparisons, mock_vs_condition1, mock_vs_condition2 and condition1_vs_condition2. The basic organizational structure of this example is key to getting an experiment with any number of baits or comparisons analyzed. Please note that deviating from this structure will likely cause 4C-Workflow to complain and stop working. Now before analyzing your own data I recommend that you follow the example below to get a basic 4C-Workflow pipeline to work.

Download Raw Fastq Files

Assuming that you have already performed git clone on the 4C-Workflow directory, you will notice the following directory organization:

1. The first thing to do is to create a folder called raw_data within 4C-Workflow/.
2. Next download the raw cd83 fastq files and extract them into raw_data/
   • Link: cd83_fastq (275 MB)
   • Although not required, it will make your life a bit easier to place raw fastq data within the raw_data/ folder.
3. Lastly, you will need the reference genome for the organism the 4C experiment was performed on, in this case Mus musculus, download and extract into the reduced_genome/ directory.
   • Link: mm10_genome (828 MB)
   • Once again although not required, placing the file into reduced_genome/ will make this process a bit easier.

Setup config.yaml

The following link will take you to the GEO page for the study. All the information you need to perform a 4C analysis on the raw cd83_1.fastq and cd83_2.fastq data is provided on this page. However to save yourself some time I will provide all the required experimental information below. Please try to enter this information into the config.yaml file before viewing the image of what the file should look like. Also please refer back to the Analysis section of this README.md if you are confused with the different parts of the config file.

Type	Value
comparison	wildtype_vs_cd43-negative
control	cd83_1
treatment	cd83_2
samples	cd83_1, cd83_2
bait_chr	chr13
bait_coord	43773612
bait_name	CD83
primary_enz	AAGCTT
primary_enz_name	hindiii
fragment_len	26
primer	CCATGACTAACTAG
species	mm
reference_genome_name	mm10

Complete Yaml

Key things to note about Complete Yaml:

• The bait name CD83 is the same for sections comparisons, samples and baits.
  • This is the case because 4C-Workflow uses this name to find all the necessary information to create the pipeline.
  • In addition, the same reason applies to why cd83_1 and cd83_2 are the same in comparisons and samples.
    • Also note how keeping the raw data and reference genome in the recommend folders shortens the path to the file location.

Before continuing please ensure that your directory structure looks exactly like the image below. If it does not then 4C-Workflow will not know where to find the necessary information.

Run

Now that 4C-Workflow has a good understanding of the experiment you can go forward and perform the analysis.

1. Open the terminal and while inside the 4C-Workflow/ folder run the following command:
   • sh runscript
2. Wait for the analysis to finish and find your results in the Output/ folder
   • Either IGV or UCSC Genome Browser can be used to view the .bedGraph and .bed files.

Note: you are working with actual raw data thus the analysis will take awhile.

Acknowledgements

This pipeline would not have been possible without the constant guidance from Ryan Dale and the awesome resources at the NIH.