The SeqAn team just finished working on their new tool Raptor, a new approach for distributing approximate searches. Find the preprint at BioRXiv. It is another tool in the DREAM framework moving it closer to completion (see also DREAM-Yara).
In comparison with similar tools like Mantis and COBS, Raptor is 12 – 144 times faster and uses up to 30 times less memory. Raptor uses winnowing minimizers to define a set of representative k-mers, an extension of the Interleaved Bloom Filters (IBF) as a set membership data structure, and probabilistic thresholding for minimizers. Our approach allows compression and a partitioning of the IBF to enable the effective use of secondary memory.