Quality control, differential gene/transcript expression and pathway analysis for bcbio RNA-seq runs
Clojure R Shell Makefile

README.md

Overview

Build Status bioconda-badge

This repository generates a Rmarkdown report from a bcbio-nextgen RNA-seq run. The report isn't a finalized analysis and definitely needs a trained person with exerpience to interpret and tweak it, but it does a lot of the annoying plumbing work when setting up an analysis. It generates a large amount of useful quality control information, runs a differential expression analysis with DESeq2 and sets up using Sleuth or doing a pathway analysis if those options are flagged on.

Installation

Conda install (UNIX only)

conda install -c bioconda bcbio-rnaseq

Manual installation

  1. Install leiningen
  2. Install pandoc (at least v1.12.3)
  3. Download the bcbio.repository
git clone https://github.com/roryk/bcbio.rnaseq
  1. Install the necessary libraries
cd bcbio.rnaseq
Rscript resources/scripts/install_libraries.R
  1. Make an executable.
make

This will make an executable in bin/bcbio-rnaseq that you can move to wherever you want and run.

Generating a bcbio-nextgen report

This produces a Rmd file with a report of a bcbio-nextgen run. If you run it without passing a formula, it will just generate a quality-control report. If you pass a formula like this:

bcbio-rnaseq summarize path-to-project-summary-yaml -f "~batch+panel"

It will run a DESeq2 analysis with the specified model and prepare reports of all pairwise comparisons for the last factor in the model. An example report is available here.

Other options you can pass are --sleuth which will set up a sleuth object to perform transcript level differential expression. Passing the --organism parameter with either mouse or human will set up functions to run clusterProfiler GSEA and gene enrichment analyses using the results of DESeq2 objects.