kcst predicts species and MLST from assembled sequences or raw reads.

kcst predicts species and multi-locus sequence type from assembled or raw reads. It does this by comparing the k-mers in the query sequences with those in the MLST profiles, picking the alleles best covered by the query.

kcst does this 'all in one go'. Instead of first predicting species in order to pick the applicable MLST scheme, it matches the query across all MLST schemes at once.

By using a k-mer mapping approach, kcst is very fast. It typically takes about a second to type one assembled genome. The binary core of kcst is khc (k-mer hit count). khc can be used standalone to rapidly compute sequence similarity between a query and a set of subject sequences.

More information: http://io.zwets.it/kcst

Assuming you are on a GNU/Linux system, and you have just cloned kcst from GitHub, here are the steps to run it:

• Install requirements

  To build kcst you need a C++ compiler and GNU make. Run c++ --version and make --version to check that you have these. Support for gzipped files requires Boost Iostreams (install the libboost-iostreams-dev package).

  Running kcst requires GNU awk, which probably is already on your system (try gawk --version), or else can be installed via the package manager.

• Build

  # Compile the khc source code
  make
  
  # Optionally run the unit tests
  make test
  

• Install

  There is no need to install khc in a specific place, only that for kcst to find it, khc must be on PATH or in the same directory as kcst.

  # Symlink bin/khc to the compiled binary src/khc
  # Note: if you cloned the repository, this link will already be present
  cd bin
  ln -sf ../src/khc
  
  # For convenience in the steps below, put kcst and khc on the PATH
  # (While in directory ./bin)
  PATH="$PWD:$PATH"
  

• Run khc

  # khc has self-contained usage instructions
  khc --help
  
  # Example: calculate coverage of ecoli.fsa by k-mers from test.fa.gz
  cd data/examples
  khc -s -k 15 -c 95 ecoli.fsa test.fa.gz
  
  # Example: pipe from stdin and sort on coverage
  zcat test.fa.gz | khc -s -k 15 -c 95 ecoli.fsa | sort -k4,4nr
  

• Run kcst

  # Construct example database with just ecoli.fsa
  cd data/examples
  ../make-kcst-db.sh -f "$PWD"  # reads file 'config', writes kcst.*
  
  # Perform MLST of test.fa against the example database
  kcst -d "$PWD" test.fa.gz
  
  # After you install the default database (see below), this suffices:
  kcst test.fa.gz
  

See the Quick Start above. kcst and khc do not need to be installed in any specific location. The only requirement is that kcst can find khc either on the PATH or in the directory where kcst itself is located.

You must provide an MLST database for kcst to work with. The simplest way to install one is by cloning and importing the database maintained by the Centre for Genomic Epidemiology (CGE) at DTU Copenhagen, as follows:

# Pick a directory to contain the downloadeded CGE database
CGE_MLST_DIR=/your/path/of/choice

# Clone the CGE git repository with the MLST database
git clone 'https://bitbucket.org/genomicepidemiology/mlst_db.git' "$CGE_MLST_DIR"

# Import the database to kcst's default MLST database directory (./data)
cd data
./make-kcst-db.sh -v -f "$CGE_MLST_DIR"  # Note default k-mer size is 15

# Check that the three database files have been created
ls kcst.*   # should give kcst.db, kcst.cfg, kcst.tsv

# Test the database: run MLST on test.fa
kcst -v examples/test.fa.gz

You can now remove CGE_MLST_DIR, or keep it around for pulling future updates:

# Update the CGE repository to the current release
CGE_MLST_DIR=/path/where/you/cloned/it
cd $CGE_MLST_DIR
git pull

# Import the database in kcst's data directory
cd $(dirname $(realpath $(command -v kcst)))/../data
./make-kcst-db.sh -f -v "$CGE_MLST_DIR"

kcst can work with multiple separate databases (see option --db), for instance with different species or k-mer sizes (the default is 15). Refer to data/README.md for information about importing MLST databases.

kcst - k-mer counting sequence typer
Copyright (C) 2018 Marco van Zwetselaar [email protected]

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.