arg_ranker

Install

pip install arg_ranker

conda install -c caozhichongchong arg_ranker

Test (download examples and use any of these commands)

arg_ranker -i example/ARGprofile_example_1.txt -m example/metadata.txt
arg_ranker -i example/ARGprofile_example_2.txt -m example/metadata.txt
arg_ranker -i test

How to use it

Prepare your ARG profile

arg_ranker is suitable for the units of ppm, gene copy per 16S or gene copy per cell

Option 1: Use our pipeline

1. Use my traits_finder to search ARGs in genomes and metagenomes (in preparation)
   Now we have both nucleotides and amino acids databases!
   https://github.com/caozhichongchong/traits_finder

2. Run
   arg_ranker -i ARG.profile.txt -m metadata.txt
arg_ranker -i ARG.profile.txt

Option 2: Run your own pipeline using our database

1. Search ARGs-OAP v1.0 database (amino acids) in your data using diamond or blast
   https://github.com/caozhichongchong/arg_ranker/tree/master/arg_ranker/data/SARG.db.fasta*

2. Format your results into example/ARGprofile_example_1.txt or example/ARGprofile_example_2.txt

3. Run
   arg_ranker -i ARG.profile.txt -m metadata.txt
arg_ranker -i ARG.profile.txt
   If you see a lot of errors saying: "ARGs in mothertable do not match with the ARGs in ARG_rank.txt.
   Please check something something in ARG.summary.cell.txt!"
   It means that the samples are placed as row names instead of colomn names (which arg_ranker expects).
   Don't worry, please try: arg_ranker -i ARG.profile.txt.t
   As we automatically transpose your table to make it work.

Option 3: Use results from ARGs-OAP v1.0 (not recommended)

1. If you have already run the ARGs-OAP v1.0 pipeline
   https://github.com/biofuture/Ublastx_stageone/archive/Ublastx_stageone.tar.gz\ https://github.com/biofuture/Ublastx_stageone/archive/Ublastx_stageone.zip

2. Check the "extracted.fa.blast6out.txt" and "meta_data_online.txt" in the output_dir

3. Run
   arg_ranker -f True -fo output_dir
arg_ranker -i formated_table.normalize_cellnumber.gene.tab -m metadata.txt

Prepare your metadata for your samples (optional)

Format your metadata of metagenomic samples into example/metadata.txt (not necessarily the same)
First column matches the sample ID in your ARG profile;
Other columns contain the metadata of your samples (such as habitat/eco-type, accession number, group...)

Introduction

ARG_ranker evaluates the risk of antibiotic resistance in metagenomes.
We designed a framework to rank the risk of ARGs based on three factors: “anthropogenic enrichment”, “mobility”, and “host pathogenicity”, informed by all available bacterial genomes, plasmids, integrons, and 850 metagenomes covering diverse global eco-habitats. The framework prioritizes 3% of ARGs in Rank I (the most at risk of dissemination among pathogens) and 0.3% of ARGs in Rank II (high potential emergence of new resistance in pathogens).

Requirement: python packages (pandas, argparse)

Requirement: a mothertable of the ARG abundance in all your samples annotated by ARGs-OAP v1.0
(see example/All_sample_cellnumber.txt).

Optimal: a table of the metadata of your samples
(see example/All_sample_metadata.txt).

Copyright

Dr. An-Ni Zhang (MIT), Prof. Eric Alm (MIT), Prof. Tong Zhang* (University of Hong Kong)

Citation

1. Zhang AN, ..., Alm EJ, Zhang T: Choosing Your Battles: Which Resistance Genes Warrant Global Action? (bioRxiv coming soon)
2. Yang Y, ..., Tiedje JM, Zhang T: ARGs-OAP: online analysis pipeline for antibiotic resistance genes detection from metagenomic data using an integrated structured ARG-database. Bioinformatics 2016.

Contact

anniz44@mit.edu or caozhichongchong@gmail.com