Skip to content

Latest commit

 

History

History
125 lines (86 loc) · 5.37 KB

tutorial.md

File metadata and controls

125 lines (86 loc) · 5.37 KB
Raw

Tutorial

This step-by-step tutorial will walk you through downloading, installing, and running MIDAS

Download and installation

MIDAS is written in Python and runs on Linux and OSX

Download the software:
git clone https://github.com/snayfach/MIDAS
read more...

Install python dependencies as needed:
python MIDAS/setup.py install
read more...

Download & unpack reference database:
https://midasdb.pollard.gladstone.org/uhgg/midas_db_v1.2.tar.gz
tar -zxvf midas_db_v1.2.tar.gz
read more...

Update your environment:
export PYTHONPATH=$PYTHONPATH:MIDAS
export PATH=$PATH:MIDAS/scripts
export MIDAS_DB=midas_db_v1.2

Download & unpack example dataset:
https://midasdb.pollard.gladstone.org/uhgg/midas_db_v1.2.tar.gz
tar -zxvf example.tar.gz

Run MIDAS

Running MIDAS can be conceptually broken down in three steps:

1. Run MIDAS per sample: run_midas.py [species, genes, snps]

  • run_midas.py species should be run prior to genes or snps
  • If you know what species you're interested in, you can skip run_midas.py species and specify them using the --species_id flag with run_midas.py genes and snps
  • run_midas.py genes and snps are independent of eachother and do not need to both be run

2. Merge results across samples: merge_midas.py [species, genes, snps]

  • Creates output matrices that can facilitate comparative analysis across samples and species/genes/snps

3. Analyze results:

  • A few scripts are provided for downstream analysis, which are listed on the Table of Contents
  • Otherwise, the output files can be analyzed and visualized using your favorite tools

1. Run MIDAS per-sample

First, create a new directory to store per-sample outputs:
mkdir midas_output

The basic command to run midas per sample is:
run_midas.py {species, genes, snps} outdir [options]

You can use the -h flag to get more info on any of the commands:
run_midas.py -h
run_midas.py species -h
run_midas.py genes -h
run_midas.py snps -h

A) Profile species abundances

run_midas.py species midas_output/sample_1 -1 example/sample_1.fq.gz
run_midas.py species midas_output/sample_2 -1 example/sample_2.fq.gz

  • Output files contain estimated species abundances for individual samples
  • This enables automatically profiling strain-level variation of all species in downstream modules
  • This step can be skipped by directly specifying species ids (--species_id) when running run_midas.py genes or snps
  • Description of output files can be found in midas_output/<sample_id>/species/readme.txt

B) Profile strain-level gene content of abundant species

run_midas.py genes midas_output/sample_1 -1 example/sample_1.fq.gz
run_midas.py genes midas_output/sample_2 -1 example/sample_2.fq.gz

  • Output files contain predicted gene +/- and estimated gene copy # for all abundant species in individual samples
  • Requires output from (1-A) unless you directly specify species using --species_id
  • Description of output files can be found in midas_output/<sample_id>/genes/readme.txt

C) Profile strain-level nucleotide variants of abundant species

run_midas.py snps midas_output/sample_1 -1 example/sample_1.fq.gz
run_midas.py snps midas_output/sample_2 -1 example/sample_2.fq.gz

  • Output files contain genome-wide nucleotide variation statistics for all abundant species in individual samples
  • Requires output from (1-A) unless you directly specify species using --species_id
  • Description of output files can be found in midas_output/<sample_id>/snps/readme.txt

2. Merge MIDAS results across samples

The basic command usage of merge scripts is:
merge_midas.py {species, genes, snps} outdir -i input -t intype [options]

  • -i indicates the sample directories output by run_midas.py
  • -t indicates the input type
  • The output of merge commands are matrix files which facilitate comparative, quantitative analyses
  • The merge command is necessary in order to perform pooled-sample, core-genome SNP calling

You can use the -h flag to get more info on any of the commands:
merge_midas.py -h
merge_midas.py species -h
merge_midas.py genes -h
merge_midas.py snps -h

A) Merge species abundance across samples

merge_midas.py species merged_species -i midas_output/sample_1,midas_output/sample_2 -t list

  • Requires that you've already run (1-A)
  • Description of output files can be found in merged_species/readme.txt

B) Merge strain-level pan-genome results across samples

merge_midas.py genes -i midas_output/sample_1,midas_output/sample_2 -t list

  • Requires that you've already run (1-B)
  • Description of output files can be found in merged_genes/<species_id>/readme.txt

C) Merge strain-level nucleotide variant results across samples

merge_midas.py snps ./merged_snps -i midas_output/sample_1,midas_output/sample_2 -t list

  • This command performs pooled-sample, core-genome SNP calling
  • Requires that you've already run (1-C)
  • Description of output files can be found in merged_snps/<species_id>/readme.txt