Skip to content

zheminzhou/SPARSE

Repository files navigation

Strain Prediction and Analysis using Representative SEquences (SPARSE)

SPARSE indexes >100,000 reference genomes in public databases in to hierarchical clusters and uses it to predict origins of metagenomic reads.

Build Status License: GPL v3 Docs Status

Installation

SPARSE runs on Unix and requires Python version 2.7 (Python 3.x supports are under development)

System modules (Ubuntu 16.04) :

  • pip
  • gfortran
  • llvm
  • libncurses5-dev
  • cmake
  • xvfb-run (for malt, optional)

3rd-party software:

  • samtools (>=1.2)
  • mash (>=1.1.1)
  • bowtie2 (>=2.3.2)
  • malt (>=0.4.0) (optional)

See requirements.txt for python module dependencies.

Installation via PIP [Suggested]

pip install meta-sparse

Installation from source codes (Ubuntu)

sudo apt-get update
sudo apt-get install gfortran llvm libncurses5-dev cmake python-pip samtools bowtie2
git clone https://github.com/zheminzhou/SPARSE
cd SPARSE/EM && make
pip install -r requirements.txt 

Updating SPARSE

You can update to latest version using PIP:

pip install --upgrade meta-sparse

If you installed SPARSE from github, move to installation directory and pull the latest version:

cd SPARSE
git pull

Quick Start

See http://sparse.readthedocs.io/en/latest/ for full documentation.

  1. Download reference database

We provide a pre-compiled database based on RefSeq (dated 19.05.2018) to download at http://enterobase.warwick.ac.uk/sparse/refseq_20180519.tar.gz . The database can be downloaded and unpacked by running:

 curl -o refseq_20180519.tar.gz http://enterobase.warwick.ac.uk/sparse/refseq_20180519.tar.gz
 tar -vxzf refseq_20180519.tar.gz

This pre-compiled database is about 350GB and contains four default mapping databases, which can be specified in the next step: representative, subpopulation, Virus, Eukaryota.

To update the database or build a costum database, please refer to the full documentation.

  1. Predict read origins

This following command will map and evaluate all reads in both fastq-files against the specified mapping databases.

sparse predict --dbname refseq_20180519 --mapDB representative,subpopulation,Virus,Eukaryota --r1 read1.fq.gz --r2 read2.fq.gz --workspace <workspace_name>

For single-end reads, only --r1 needs to be specified. All output files are stored in the respective workspace.

  1. Create a report
sparse report <workspace_name>

The report will be stored in <workspace_name>/profile.txt

  1. Extract reference specific reads

The following command extracts all reads specific to the provided reference ids, which can be found in the output of step 2.

sparse extract --dbname refseq_20171014 --workspace <workspace_name> --ref_id <comma delimited indices>

Citation

SPARSE is published as a conference proceeding in "Research in Computational Molecular Biology".

Zhemin Zhou, Nina Luhmann, Nabil-Fareed Alikhan, Christopher Quince, Mark Achtman, 'Accurate Reconstruction of Microbial Strains from Metagenomic Sequencing Using Representative Reference Genomes' RECOMB 2018: Research in Computational Molecular Biology pp 225-240. doi: https://doi.org/10.1007/978-3-319-89929-9_15

A preprint version of the manuscript is also accessible in bioRxiv: https://doi.org/10.1101/215707