Barapost command line toolkit is designed for binning (i.e. separation into different files) FASTA, FASTQ and FAST5 files according to taxonomic annotation. Taxonomic annotation is implemented as finding the most similar reference sequence in a nucleotide database: remotely using NCBI BLAST web service or on a local machine with BLAST+ toolkit.

• Demultiplexing whole genome sequencing reads (basically, "long" reads) without barcoding. This demultiplexing is based on taxonomic annotation, therefore organisms of interest should be distant (having in mind naive classification algorithm and lateral gene transfer). Usually, it is enough for organisms to belong to different genera.
• Genome assembly: Barapost can detect and remove contigs assembled from cross-talks. You might have seen such contigs: they are short and have low coverage, they don't belong to genome of interest and should be removed.
• Comparing different sets of contigs of the same genome (see Example 7 on barapost-local page).
• I dare to surmise this list isn't complete :-)

You can get "Barapost" in the following ways:

Way 1: go to terminal and run git clone https://github.com/masikol/barapost.git

Way 2: download ZIP archive (green button "Code" at the top right of this page. Then "Download ZIP".

After that, you can find executable Python scripts in directory barapost-master/barapost. Scripts can be run in place.

Or you can add this directory to your PATH variable so that you won't need to enter full script's path to run it -- just it's name, for example:

barapost-prober.py some_reads.fastq

instead of

/home/user/some/more/long/useless/path/barapost/barapost/barapost-prober.py some_reads.fastq

If you are not sure how to do this, following links can help you:

• Editing PATH on Linux
• Editing PATH on Windows
• Editing PATH on MacOS

1. barapost-prober.py -- this script submits several sequences (i.e. only a part of your data set) to NCBI BLAST server in order to determine what taxons are "present" in data set. "barapost-prober.py" saves accession numbers of best hit(s) of each submitted input sequence. Processing all sequences in this way takes too much time, what leads us to "barapost-local.py".

2. barapost-local.py -- this script firstly downloads best hits "discovered" by "barapost-prober.py" from GenBank, then composes a database from downloaded reference sequences on local machine and finally classifies the major part of data using created database. "barapost-local.py" creates a database and "BLASTs" input sequences with "BLAST+" toolkit.

3. barapost-binning.py -- this script bins (divides into separate files) nucleotide sequences according to results of "barapost-prober.py" and/or "barapost-local.py"

You can find detailed information about Barapost at Barapost Wiki.

You can find paper describing the toolkit together with benchmark at https://doi.org/10.1109/TCBB.2020.3009780. If you use Barapost in your research, it would be great if you could cite us.