snakefiles

This repository has Snakefiles for common RNA-seq data analysis workflows. Please feel free to copy them and modify them to suit your needs.

Getting started

If you are new to [Snakemake], you might like to start by walking through my [tutorial for beginners][beginners]. Next, have a look at Johannes Koster's [introductory slides][slides], [tutorial], [documentation], and [FAQ].

Quick start:

# Copy the files
git clone https://github.com/slowkow/snakefiles.git
Go to the kallisto directory
cd snakefiles/kallisto
Run snakemake
snakemake

Data

This repository includes 6 FASTQ files in [data/fastq/][fastq] to illustrate the usage of each of the RNA-seq workflows.

• Sample1

- Sample1.R1.fastq.gz has the first mates of sequenced fragments. - Sample1.R2.fastq.gz has the second mates of sequenced fragments.

• Sample2

- Sample2.L1.R1.fastq.gz - Sample2.L2.R1.fastq.gz - The first mate reads (R1), split across two files (L1 and L2). Some software such as STAR requires these reads to be merged into one file. - Sample2.L1.R2.fastq.gz - Sample2.L2.R2.fastq.gz - Likewise, the second mate reads (R2) are also split across two files (L1 and L2). To make matters worse, Sample2.L2.R2.fastq.gz has only 2000 reads, whereas Sample2.L2.R1.fastq.gz has 2500 reads. The Snakefiles in this repository can handle this without any problems.

[fastq]: https://github.com/slowkow/snakefiles/tree/master/data/fastq

Scripts

• [makesamples.py][makesamples] creates the [samples.json][samples] file.
• [bsub.py][bsub] receives job scripts from Snakemake and automatically

submits them to an appropriate LSF queue based on job requirements.

[makesamples]: https://github.com/slowkow/snakefiles/tree/master/makesamples.py [samples]: https://github.com/slowkow/snakefiles/tree/master/samples.json [bsub]: https://github.com/slowkow/snakefiles/tree/master/bsub.py

RNA-seq workflows

[kallisto/][1]

[1]: https://github.com/slowkow/snakefiles/tree/master/kallisto

Quantify gene isoform expression in transcripts per million (TPM) with [kallisto] and collate outputs from multiple samples into one file.

[star_express/][3]

[3]: https://github.com/slowkow/snakefiles/tree/master/star_express

Execute a multi-sample 2-pass [STAR] alignment, sharing the splice junctions across samples. Count fragments per gene and fragments per splice site. Also produce a BAM file with coordinates relative to transcripts. Quantify transcripts in TPM with [eXpress]. Collate outputs from multiple samples.

Contributing

Please [submit an issue][issues] to report bugs or ask questions.

Please contribute bug fixes or new features with a [pull request][pull] to this repository.

[issues]: https://github.com/slowkow/snakefiles/issues [pull]: https://help.github.com/articles/using-pull-requests/

[kallisto]: https://github.com/pachterlab/kallisto [STAR]: https://github.com/alexdobin/STAR [eXpress]: http://bio.math.berkeley.edu/eXpress/overview.html [cc0]: https://creativecommons.org/publicdomain/zero/1.0/

[beginners]: http://slowkow.com/notes/snakemake-tutorial/

[Snakemake]: https://bitbucket.org/snakemake/snakemake/wiki/Home [slides]: http://slides.com/johanneskoester/deck-1 [tutorial]: http://htmlpreview.github.io/?https://bitbucket.org/snakemake/snakemake/raw/master/snakemake-tutorial.html [documentation]: https://bitbucket.org/snakemake/snakemake/wiki/Documentation [FAQ]: https://bitbucket.org/snakemake/snakemake/wiki/FAQ