LINKS 1.5 (Apr 24, 2015)
Bloom filter implementation for reduced memory usage and faster execution.
For additional information about this project, please visit the overview page .
Available downloads
Release Notes
| State | Final release |
|---|---|
| License | GPL |
| Release Manager | Rene Warren |
LINKS now uses a Bloom filter to limit hashing paired k-mers to only those found in sequences to re-scaffold. This feature decreases RAM usage by over 60%, while the run time is nearly unchanged (tests on E. coli, S. typhi, S. cerevisiae genomes). But when ran iteratively, users can re-use Bloom filters with the -r option, which results in faster* run times compared to v1.3 and earlier. *Approx. half the time needed on S. cerevisiae assemblies (eg. ~3h vs. 6h. for 30 iterations, with settings described in paper). Note that Bloom filter construction time may be long for large genomes such as human (clocked at 6h for k=26 on system below) and 20-Gb white spruce, and the users may want to build the filter ahead of LINKS using the utility provided in the ./tools folder.
LINKS v1.5 and unreleased v1.4 (Bloom filter implementation) benchmarks compared to v1.3
