snakemake-workflows/transcriptome-differential-expression

This project tries to re-animate the project from Oxford Nanopore. CURRENTLY IT IS NOT IN A WORKING STATE. Please see their current nextflow implementation as a reference wf-transcriptomes, which contains functionality for differential expression.

Pipeline for differential gene expression (DGE) and differential transcript usage (DTU) analysis using long reads

This pipeline uses snakemake, minimap2, salmon, edgeR, DEXSeq and stageR to automate simple differential gene expression and differential transcript usage workflows on long read data.

If you have paired samples (e.g for example treated and untreated samples from the same individuals) use the paired_dge_dtu branch.

Getting Started

Input

The input files and parameters are specified in config.yml:

transcriptome - the input transcriptome.
annotation - the input annotation in GFF format.
condition_a_identifier - a string identifiying the first trait.
condition_b_samples - a string identifiying the second trait.

Output

alignments/*.bam - unsorted transcriptome alignments (input to salmon).
alignments_sorted/*.bam - sorted and indexed transcriptome alignments.
counts - counts generated by salmon.
merged/all_counts.tsv - the transcript count table including all samples.
merged/all_counts_filtered.tsv - the transcript count table including all samples after filtering.
merged//all_gene_counts.tsv - the gene count table including all samples.
de_analysis/coldata.tsv - the condition table used to build model matrix.
de_analysis/de_params.tsv - analysis parameters generated from config.yml.
de_analysis/results_dge.tsv and de_analysis/results_dge.pdf- results of edgeR differential gene expression analysis.
de_analysis/results_dtu_gene.tsv, de_analysis/results_dtu_transcript.tsv and de_analysis/results_dtu.pdf - results of differential transcript usage by DEXSeq.
de_analysis/results_dtu_stageR.tsv - results of the stageR analysis of the DEXSeq output.
de_analysis/dtu_plots.pdf - DTU results plot based on the stageR results and filtered counts.

Dependencies

miniconda - install it according to the instructions.
snakemake install using conda.
pandas - install using conda.
The rest of the dependencies are automatically installed using the conda feature of snakemake.

Layout

README.md
Snakefile - master snakefile
config.yml - YAML configuration file
snakelib/ - snakefiles collection included by the master snakefile
lib/ - python files included by analysis scripts and snakefiles
scripts/ - analysis scripts
data/ - input data needed by pipeline - use with caution to avoid bloated repo
results/ - pipeline results to be commited - use with caution to avoid bloated repo

Installation

Clone the repository:

git clone https://github.com/snakemake-workflows/transriptome-differential-expression

Usage

Edit config.yml to set the input datasets and parameters then issue:

On a server, e.g.:

snakemake --use-conda -j <num_cores> all

On a cluster, e.g.

snakemake --slurm --default-resources slurm_account=<your slurm account> slurm_partition=<your clusters default partition> -j <unlimited or lower> --configfile ./envs/<your config yaml> --workflow-profile ./profile/ --snakefile <path to Snakefile> --directory <desired working directory>

Note, that the profile offers a template cluster configuration - it needs adjusting for particular clusters. Contributions of particular configurations are welcome!

Help

Licence and Copyright

This Source Code Form is subject to the terms of the Mozilla Public License, v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain one at http://mozilla.org/MPL/2.0/.

References and Supporting Information

This worflow is largely based on the approach described in the following paper:

Love MI, Soneson C and Patro R. Swimming downstream: statistical analysis of differential transcript usage following Salmon quantification. F1000Research 2018, 7:952 (doi: 10.12688/f1000research.15398.3)

snakemake-workflows / transcriptome-differential-expression