/smsk_khmer_trinity

Transcriptome assembly via digital normalization with khmer and Trinity

Primary LanguagePythonMIT LicenseMIT

smsk_khmer_trinity: a simple workflow for transcriptome assembly

Build Status

1. Description

This is a workflow for de novo transcriptome assembly with Illumina reads. It

  1. Trims reads with Trimmomatic

  2. Performs digital normalization with khmer

  3. Assembles with trinity

2. First steps

Just follow what is inside the .travis.yml

  1. Install conda

  2. Clone this repo

  3. Add your samples to config.yaml

  4. Run snakemake: snakemake --use-conda -j

3. File organization

The hierarchy of the folder is the one described in A Quick Guide to Organizing Computational Biology Projects:

smsk_khmer_trinity
├── bin: binaries, scripts and environment files.
├── data: raw data, hopefully links to backup data.
├── README.md - This
├── results: processed data.
|    ├── raw: links to raw data
|    ├── qc: processed reads with trimmomatic
|    ├── diginorm: digital normalization
|    ├── assembly: Trinity output
|    ├── filtering: TPM per loci filtering
|    ├── tissue: per sample quantification
|    └── transrate: assembly and filtering statistics
└── src: additional source code, tarballs, etc.

4. Analyzing your data

"Just" edit the config.yaml with the paths to your fastq files and change parameters. In the section diginorm_params \ max_table_size type 4e9 because it's anoyingly slow to do tests with 16Gb of RAM.

Also raise Trinity's maximum memory usage if you need it.

Links, References and Bibliography