k-SLAM: Accurate and ultra-fast taxonomic classification and gene identification for large metagenomic datasets

k-SLAM is a highly e cient algorithm for the characterisa- tion of metagenomic data. Unlike other ultra-fast metage- nomic classi ers, full sequence alignment is performed allow- ing for gene identi cation and variant calling in addition to accurate taxonomic classi cation. A k -mer based method provides greater taxonomic accuracy than other classi ers and a three orders of magnitude speed increase over align- ment based approaches. The use of alignments to nd vari- ants and genes along with their taxonomic origins enables novel strains to be characterised. k-SLAM's speed allows a full taxonomic classi cation and gene identi cation to be tractable on modern large datasets. A pseudo-assembly method is used to increase classi cation accuracy by up to 40% for species which have high sequence homology within their genus