Metagenomic Read Mapping


CuCLARK is a metagenomic classifier for CUDA-enabled GPUs, based on CLARK The program comes in two variants: CuCLARK and CuCLARK-l. CuCLARK is designed for workstations which can provide enough RAM to fit large databases (like all bacterial genomes in NCBI/RefSeq) and an arbitrary number of CUDA-enabled GPUs. CuCLARK-l uses a much smaller database and can be run on computers with 4 GB of RAM and a CUDA-enabled GPU with at least 1 GB of memory.


MetaCache is a classification system for mapping short reads from metagenomic samples to their most likely taxon of origin. MetaCache aims to reduce the memory requirement usually associated with k-mer based methods while retaining their speed. It uses locality sensitive hashing to quickly identify candidate regions within one or multiple reference genomes. A read is then classified based on the similarity to those regions.