RNA-Seq analysis standard pipeline for paired-end data

⚠️ Warning

To use this pipeline use the Fork option, on the top right of the repo.

Description

This pipeline uses the Snakemake workflow manager for PE RNA sequencing data processing.

Workflow

This is the current workflow for the snakemake rules.

Getting started

1. Fork the pipeline to create your own project repo.

2. Create the config/sample_manifest.tsv using the following structure:

   1. Identifier: Facility ID

   2. Condition: Condition (treated/untreated, fed/starved, etc.)

   3. Bio_rep: Biological replicate number

   4. Target_genome: ENCODE target genome for sequencing (supported options: M25, M32, 19, 38)

   5. Sequencer: Sequencer to define —2colour parameter for the trimming (supported options: HiSeq4000, NovaSeq, NextSeq500)

   6. Fastq_handle: Particular handle useful for raw .fastq.gz files selection using the name from the facility (number, extension, etc.)

      Sample manifest example available here.
      
      

3. Copy raw sequencing data. For this there are 2 input options (data with 2 sequencing lanes, unique lane/merged files).

   1. For fastq.gz files from 2 sequencing lanes. Copy all your fastq.gz to resources/fastq_seq/raw, with no additional subfolder and keeping the original facility names (see bellow for the usual facility filenames). Your folder should look like this:

      resources/fastq_seq/raw
          ├── {Identifier}_S{Fastq_handle}_L001_R1_001.fastq.gz
          ├── {Identifier}_S{Fastq_handle}_L002_R1_001.fastq.gz
          ├── {Identifier}_S{Fastq_handle}_L001_R2_001.fastq.gz
          └── {Identifier}_S{Fastq_handle}_L002_R2_001.fastq.gz
      

   As a result you should have 4 fastq.gz files per sample (2 reads with 2 lanes each).

   ⚠️ Warning

   Note that the {wildcards} correspond to the columns on the config/sample_manifest.tsv .

   2. For fastq.gz files from merged lanes or unique. Manually copy your fastq.gz to resources/fastq_seq/merged, with a subfolder for each sample, named using the following structure:

      resources/fastq_seq/merged
          ├── {condition}_Bio-rep_{Bio_rep}
          │   ├── {condition}_Bio-rep_{Bio_rep}_R1.fq.gz
          │   └── {condition}_Bio-rep_{Bio_rep}_R2.fq.gz
          │
          └── {condition}_Bio-rep_{Bio_rep}
              ├── {condition}_Bio-rep_{Bio_rep}_R1.fq.gz
              └── {condition}_Bio-rep_{Bio_rep}_R2.fq.gz
      

   As a result you should have 2 fq.gz files per sample (2 reads each).

   ⚠️ Warning

   Note that the {wildcards} correspond to the columns on the config/sample_manifest.tsv . And the extension is fq.gz instead of fastq.gz .

4. Change processing parameters on your config/config.yaml.

5. Run the pipeline.

   Your pipeline should look like this:

   <YOUR PROJECT FOLDER>
       ├── config
       │   ├── config.yaml
       │   └── sample_manifest.tsv
       ├── profiles
       |   └── profile
       ├── resources
       |   └── fastq_seq
       |       └── raw or merged
       └── workflow
   

   Use the following command on your project folder:

       snakemake --profile profiles/<YOUR_PROFILE> --use-conda -j<N_JOBS>
   

For more options checkout the --help

Important notes

• This pipeline only works with paired-end Illumnia sequencing reads

• The original pipeline has a merge_fq_lanes rule adapted for particular input fastq.gz filenames, for more information check step 3 on Getting started.

• The default configurations for this workflow are suitable to run snakemake on the Helmholtz-Munich HPC and using conda to deal with software dependencies. If one wishes to run the pipeline on a different computing platform, the profiles need to be adapted accordingly.

• The pipeline only supports GENCODE genomes as reference, choose the right GENCODE Realease as value for the Target_genome entry in your config/sample_manifest.tsv. These are the equivalent realeases for GENCODE and UCSC databases:

  Organism	GENCODE Realease	GENCODE Genome	UCSC Genome
  Mouse	M25	GRCm38	mm10
  Mouse	M32 (Latest)	GRCm38	mm39
  Human	19	GRCh37	h19
  Human	38	GRCh38	h38

Authors and acknowledgment

Paulina Rosales-Becerra, Kevin Brokers, Saulius Lukauskas & Robert Schneider

Contact

paulina.rosales@helmholtz-muenchen.de

Project status

Actual version works just fine :)