- k-SLAM is a program for ultra-fast and accurate alignment based metagenomic analysis of large sets of high-throughput sequence data.
- k-SLAM uses a k-mer based technique to rapidly find alignments between reads and genomes which are then validated using the Smith-Waterman algorithm.
- Alignments are chained together into a pseudo-assembly to increase specificity.
- Taxonomy is inferred using a Lowest Common Ancestor technique.
- Output is presented as a Krona visualisation of per-read abundance.
- Genes and variants are inferred from the alignments and output in SAM format.
- k-SLAM is fast and highly parallelisable, with speeds of 5 million reads per minute on 150bp paired end data.
This software was written by David Ainsworth under the supervision of Sarah Butcher, Michael Sternberg and Come Raczy. This work was funded by the Biotechnology and Biological Sciences Research Council CASE Studentship grant BB/I01585X/1 with Illumina Cambridge.
k-SLAM is freely available under the GNU Affero General Public License.