This project is deprecated. Please see our newer 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.control_samples
- a dictionary with control sample names and paths to the fastq files.treated_samples
- a dictionary with treated sample names and paths to the fastq files.
Output
alignments/*.bam
- unsorted transcriptome alignments (input tosalmon
).alignments_sorted/*.bam
- sorted and indexed transcriptome alignments.counts
- counts generated bysalmon
.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 fromconfig.yml
.de_analysis/results_dge.tsv
andde_analysis/results_dge.pdf
- results ofedgeR
differential gene expression analysis.de_analysis/results_dtu_gene.tsv
,de_analysis/results_dtu_transcript.tsv
andde_analysis/results_dtu.pdf
- results of differential transcript usage byDEXSeq
.de_analysis/results_dtu_stageR.tsv
- results of thestageR
analysis of theDEXSeq
output.de_analysis/dtu_plots.pdf
- DTU results plot based on thestageR
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 ofsnakemake
.
Layout
README.md
Snakefile
- master snakefileconfig.yml
- YAML configuration filesnakelib/
- snakefiles collection included by the master snakefilelib/
- python files included by analysis scripts and snakefilesscripts/
- analysis scriptsdata/
- input data needed by pipeline - use with caution to avoid bloated reporesults/
- pipeline results to be commited - use with caution to avoid bloated repo
Installation
Clone the repository:
git clone https://github.com/nanoporetech/pipeline-transcriptome-de.git
Usage
Edit config.yml
to set the input datasets and parameters then issue:
snakemake --use-conda -j <num_cores> all
Help
Licence and Copyright
(c) 2018 Oxford Nanopore Technologies Ltd.
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 pipeline 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)
See the post announcing the tool at the Oxford Nanopore Technologies community here.