Validation of ASaiM framework on HMP Mock Community samples

We test ASaiM framework on mock samples from HMP which have a controlled community content. The results were compared to EBI metagenomic pipeline results.

Requirements

• conda
• ASaiM framework with its custom Galaxy instance launched and populated with tools and databases. You can also check ASaiM documentation for further information.

Once conda is installed, create the conda environment (with all required dependencies):

$ conda config --add channels biocore
$ conda config --add channels bioconda
$ conda env create -f conda_env.yml

Usage

Activate conda environment:

$ source activate hmp_mock

Generate an API key corresponding to your account on the custom Galaxy instance (in User menu, on top panel) of ASaiM framework, and fill the config.yml file:

asaim_galaxy_instance_url: "http://0.0.0.0:8080/"
api_key_on_asaim_galaxy_instance: "apikey"

Launch analyses:

$ ./src/launch_hmp_mock_analyses.sh

This script will

• Get the input datasets, EBI result data and format them
• Get reference genomes, reference rRNA sequences and mapped input datasets on them
• Launch ASaiM workflow on both datasets (this task takes several hours). You can visualize workflows running by browsing ASaiM Galaxy instance

Export results and analyze them:

$ ./src/export_analyze_hmp_mock_results.sh

This script will

• Export ASaiM workflow outputs (when the workflows are done) et generate all graphics
• Concatenate results (EBI one and ASaiM one) to compare them

Report

A report of this analysis is available in doc/report directory.

To generate the PDF from the markdown file (requiring PANDOC):

$ pandoc -s doc/report.md -t latex -o doc/report.pdf --bibliography="doc/references.bib" -N