Differential Gene Expression with Snakemake

This Snakemake workflow uses Fastp to QC the reads, Salmon for transcript quantification, Tximport to convert from transcript abundance to gene abundance, and DEseq2 for differential gene expression.

INSTALLATION

• Clone the repository:
  • git clone https://github.com/davidecarlson/Snakemake_diff_gene_expr.git
• Use conda to install create a new environment with all the necessary software from the provided yaml file:
  • conda env create -f environment.yml

SETUP

• Edit the example_config.yaml file to provide information regarding your data and experiment:

  • path to the RNA-Seq reads
  • path to save the results
  • information regarding the experimental conditions and samples being used
  • number of threads to use
  • species species ("human", "mouse", or "neither")
  • path to your Salmon index, which must be built ahead of time
    • see here for instructions on building the index
  • path to gtf file of gene annotations, if "neither" was chosen as your species
  • path to tab-separated file with list of samples and associated experimental condtions (e.g., "samples.txt")
  • name of the condition from "samples.txt" that should act as the control
  • whether or not input reads are gzipped
  • whether or not input reads are paired

• Edit the samples.txt file to include a row for each sample, including sample name, run name (can be the same as sample name if there is only one run per sample), and experimental condition for that sample

USAGE

• conda activate snakemake_DGE
• snakemake --cores <# of threads * # of processes to run currently> -s <path to snakefile>

All output files and logs will be saved in the directory specificed in your config file

TO-DO

• Add more automation so that fewer files need to be manually edited prior to usage