A hybrid approach designed to achieve extremely rapid phylogenetic classification of metagenomic sequences. TWARIT makes use of both 'sequence similarity' and 'sequence composition' to achieve the final taxonomic binning. First, TWARIT utilizes Burrows-Wheeler Aligner's (BWA) short-read alignment to classify metagenomic sequences originating from genomes that already exist in the reference database. In the next step, sequences originating from novel organisms are classified by the lexicographic sorting of their unique compositional signatures along with those of reference sequences. Both these steps are extremely rapid making TWARIT as one of the fastest binning methods. At the same time, TWARIT achieves reasonably good levels of binning accuracy and specificity.