BASTA

BAsic Sequence Taxonomy Annotation

As the name implies, BASTA assigns taxonomies to sequences or groups of sequences based on the Last Common Ancestor (LCA) of a number of best hits. BASTA can be customised to run on any kind of tabular output (default = blast -outfmt 6) as long as the input file provides values for e-value, percent identity and alignment length. Taxonomies are inferred from NCBI taxonomies based on a 7 level taxonomy.

For detailed usage and installation instructions please visit https://github.com/timkahlke/BASTA/wiki

Citing BASTA

If you like BASTA and use it for publications please cite it as "Kahlke T and Ralph PJ (2018), BASTA–Taxonomic classification of sequences and sequence bins using Last Common Ancestor estimations. Meth. Ecol. Evol. doi:10.1111/2041‐210X.13095"

Requirements

BASTA >= v1.4 is written in python3 and requires the following dependencies:

leveldb
plyvel
python-wget
krona

Installation

Installing BASTA v1.4 in conda environment (Recommended)

It's recommended to install python in a conda environment as follows:

# Create a conda ennvironment called basta_py3
conda create -n basta_py3 python=3

# Activate the conda environment
conda activate basta_py3

# Download and install dependencies
conda install -c bioconda -c conda-forge leveldb plyvel krona python-wget

# Clone the github directory
git clone https://github.com/timkahlke/BASTA.git
cd BASTA
python setup.py install

For installation without conda see installation instructions on the wiki (https://github.com/timkahlke/BASTA/wiki).

Installing legacy versions (python2)

Old python2 versions can be installed completely using conda:

conda install -c bioconda -c bnoon -c timkahlke basta

The above command will install the required dependencies leveldb and wget as well as the required python libraries.

BASTA installation has been tested for MacOSX 10.11 and Ubuntu 14.01 and 16.01 using miniconda2.

Quick start

Inital Setup

# set up NCBI taxonomy database
basta taxonomy

# download and set up genbank and uniprot mappings
# NOTE: this might not be needed for you. See Wiki for details
basta download gb
basta download prot

Conda default data directory

Per default the taxonomy data as well as mapping databasese are stored in $HOME/.basta.

Running BASTA

# Infer one LCA for each query sequence of blast against uniprot
basta sequence BLAST_OUTPUT_FILE BASTA_OUTPUT_FILE prot

# Infer one LCA for the complete blast output file
basta single BLAST_OUTPUT_FILE prot

# Infer one LCA for each blast output file in a given directory
basta multiple BLAST_OUTPUT_DIRECTORY BASTA_OUTPUT_FILE prot

Last Common Ancestor algorithm

BASTA supports taxonomic estimation based on a percentage of best hits using the -p flag.

100% = all

If set to 100% (default) BASTA reads a given number of best hits for each query sequence and returns the LCA of all sequences.

51 - 99% = majority classification

If set to a value between 51 and 99 BASTA returns the taxonomy that is shared by at least the given percentage of hits. This gives the user the potential to build a majority taxonomy instead of including all best hits in the BASTA result.

Additional scripts

basta2krona.py

This creates a krona plot (html file) that can be opened in your browser from a basta annotation output file.

basta2krona BASTA_OUTPUT_FILE KRONA_HTML_FILE

filter_basta_fasta.py

This script can be used to filter a given fasta file based on BASTA annotations.

filter_basta_fasta.py [options] FASTA_FILE FILTERED_OUTPUT_FILE NAME_OF_TAXON BASTA_FILE

timkahlke / BASTA

readme