Amplicons to Global Gene (A2G²)

This program implements the progressive algorithm to align a large set of amplicons to a reference gene consensus, or a large set of sequences to an amplicon consensus, based on a reference consensus. This program makes use of traditional multiple aligners such as MAFFT (default), and muscle, and can be extended to other aligners.

Problem

Some taxonomic assignment software require a set of align sequences, both in the query as in the reference. Projects such as those using environmental DNA (eDNA) or trying to assess wide diversity using metagenomics often have a hard time creating such alignments, because of memory and computational restrictions. Another observation is that massive alignments often introduce more gaps in the sequences, and force alignment of segments that should not align in that region. Here is where A2G² will use a global to local alignment to avoid such issues, and retained the ungapped alignment of the amplicons.

Installing

A2G² can be install it through PyPi, by:

pip install A2G

Alternatively, you can clone this repository, and install it using setuptools:

git clone https://github.com/jshleap/A2G.git
cd A2G
python3 setup.py install

Basic usage

A2G² will give you help by:

A2G -h

this should give you something like this:

A2G version: 2020.0.1
Copyright 2020 Jose Sergio Hleap

usage: A2G [-h] [--cpus CPUS] [--nowrite] [--out_prefix OUT_PREFIX]
           [--remove_duplicates]
           global_consensus local_consensus fasta

positional arguments:
  global_consensus      Sequence consensus of the global region, e.g. full COI
  local_consensus       Sequence consensus of the local region, e.g. Leray
                        fragment
  fasta                 fasta file with the focal sequences

optional arguments:
  -h, --help            show this help message and exit
  --cpus CPUS           number of cpus to use
  --nowrite             return string instead of writing
  --out_prefix OUT_PREFIX
                        Prefix of outputs
  --remove_duplicates   Keep or remove duplicated sequences

Then to run it, you can simply type:

A2G global_target local_target query_file --cpus 10 --out_prefix prefix --remove_duplicates

With this command, you will use the global_target as the overall region, the local_target as the amplicon reference sequence to anchor the query sequences, and query_file contains your query sequences. Those are the required arguments. The optional arguments allow you to control the execution. --cpus allow you to provide the number of cpus to use. In the example, up to 10 cpus will be used. --out_prefixchange the prefix of the outputs generated. Finally, the --remove_duplicates option will retain only unique sequences.

If the no_write option is used, A2G² will output the alignment to standard out, and other info to standard error. If you would like to pipe only the alignment, you can redirect the standard error to a null device:

A2G global_target local_target query_file --cpus 10 --out_prefix prefix --no_write 2> /dev/null