Table of Contents

Description
Installation
Usage

Description

This repository contains a demonstration of a basic bulk RNA-seq pipeline in two forms, Snakemake (in the workflow/ directory) and as a bash script (in the bash_workflow/ directory).

The two workflows perform the same analysis, but the Snakemake pipeline is much more robust and recommended, though not perfect just yet. This Snakemake pipeline was created as part of a tutorial, and thus refrains from employing some of the full/more complex utilities of Snakemake . Still, it is a maked improvement over the bash workflow.

A couple of advantages of the Snakemake version.

• Stops at an error, removes incomplete files automatically

• Can rerun/restart a pipeline after an error and only runs steps that were not completely finished

• Can run to a certain point within the pipeline without editing the pipeline

• Can add new samples and rerun the pipeline so that it only runs the steps necessary for processing of the new samples and integration of those new data with other sample data, if necessary

• Can generalize the analysis

• Do not have to pre-install all software

• Can have temporary files removed automatically

• With the use of a config.yaml file, need only to edit a single file to run a new analysis

• And more...

Steps performed

The Snakemake workflow structure

$ tree workflow
workflow
├── Snakefile
├── envs
│ ├── fastqc_env.yml
│ ├── star_env.yml
│ ├── subread_env.yml
│ └── trim_galore_env.yml
├── resources
│ ├── config.yaml
│ └── metadata.txt
└── scripts
 └── DESeq2.R

A brief description of the files within the workflow

config.yaml File Field Descriptions and Example

Example config.yaml

out_dir: '/Users/csifuentes/Documents/outData'
input_dir: fastq: '/Users/csifuentes/Documents/testData/simulated_reads/fastq' genome_fasta: '/Users/csifuentes/Documents/testData/index/' genome_gtf: '/Users/csifuentes/Documents/testData/index/Homo_sapiens.GRCh38.103.gtf'
read_info: fastq_suffix: 'fastq' r1_suffix: '_1'
thread_info: fastqc: 4 align: 4 counts: 4
trim_info: length: 20 quality: 30
count_info: strand: 0
diffex_info: prefix: 'WT_vs_KO' metadata: '/Users/csifuentes/Documents/repo/snakemake-demo-rnaseq/workflow/resources/metadata.txt'

Key Assumptions

• Files are gzipped, ending with .gz .

• Paired-end reads.

• Only 2 groups to compare (WT vs KO, or T0 vs T24, etc.)

Table of Contents

Installation:

1. Install miniconda using the appropriate directions from here: https://docs.conda.io/projects/conda/en/latest/user-guide/install/index.html

2. After closing and re-opening the terminal, the conda base environment should be enabled. To check this, make sure you see (base) to the left of your user name in the terminal prompt.

(base) user: $

1. Install mamba using conda in your base environment.

conda install -n base conda-forge mamba

1. Install Snakemake using conda / mamba . I install it into it's own environment, named snakemake_env here.

conda activate base
mamba create -c conda-forge -c bioconda -n snakemake_env snakemake

1. Go to https://github.com/cjsifuen/Basic-Bulk-RNA-seq-Workflow. Click the green Code button on the top-right side of the page, and select Download ZIP , shown in the purple box below.

1. Move the downloaded file Basic-Bulk-RNA-seq-Workflow-main.zip to where you want to work, and unzip (right-click and extract or unzip, or double-clicking usually works).

2. If you have not installed R, install it from here: https://rstudio-education.github.io/hopr/starting.html#how-to-download-and-install-r. You can check for an R installation by opening the terminal and typing R --version , followed by the enter key. If R is installed the version number and other information will be shown.

$ R --version
R version 4.0.3 (2020-10-10) -- "Bunny-Wunnies Freak Out"
Copyright (C) 2020 The R Foundation for Statistical Computing
Platform: x86_64-apple-darwin17.0 (64-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under the terms of the
GNU General Public License versions 2 or 3.
For more information about these matters see
https://www.gnu.org/licenses/.

1. Install DESeq2 in R in your snakemake_env by first activating the snakemake_env . Then follow the installation instructions here: https://bioconductor.org/packages/release/bioc/html/DESeq2.html.

# activate the environment
conda activate snakemake_env

# open R
R

# install DESeq2
if (!requireNamespace("BiocManager", quietly = TRUE))
 install.packages("BiocManager")
BiocManager::install("DESeq2")

1. Exit R by typing q() , followed by enter. Select n for no. Everything should be ready.

Table of Contents

Usage:

1. Edit the config.yaml file, described above in description . Note the key assumptions being made about these data/files.

2. Edit the metadata.txt file.

3. Open the terminal and activate the snakemake_env

conda activate snakemake_env

1. Change into the directory containing the unzipped repository. If I downloaded the repository to my Downloads directory and unzipped it there, I would type

cd /Users/csifuentes/Downloads/Basic-Bulk-RNA-seq-Workflow-main

1. Test the pipeline to make sure everything looks okay.

snakemake -n

Should produce the following (as well as a larger block of text). This shows the number of jobs that will be run for each rule. So, for 6 samples, we have 12 fastq files (since they are paried). Thus, I should see 12 fastqc jobs that need to be performed. I should also see 6 star jobs for alignment. I see these below. Everything looks good so I can continue to run the analyis.

Job counts:	count	jobs	1	all	1	clean_counts	1	deseq2	12	fastqc	1	featureCounts	6	star	6	trim_galore	28
This was a dry-run (flag -n). The order of jobs does not reflect the order of execution.

Check that the number of fastqc run

1. Run the pipeline. Some useful parameters are shown below.

snakemake --use-conda --conda-frontend mamba --cores 8 -p 2>&1 | tee log.file

Table of Contents