DSK is a k-mer counting software, similar to Jellyfish. Jellyfish is very fast but limited to large-memory servers and k ≤ 32. In contrast, DSK supports large values of k, and runs with (almost-)arbitrarily low memory usage and reasonably low temporary disk usage. DSK can count k-mers of large Illumina datasets on laptops and desktop computers.
Reference
G. Rizk, D. Lavenier, R. Chikhi. (2013) DSK: k-mer counting with very low memory usage, Bioinformatics, 29(5):652-3 [PDF]
Ready-to-use executable
DSK is available as a binary for immediate use on Linux and MacOSX platforms, with the following requirements:
MacOS-X 10.8 or above. (Intel 64bit processors) |
Download | |
Linux running on Intel or AMD 64bit processors. (kernel 2.6.32 or above, GLIBCXX_3.4.13 or above) |
Download |
For all other platforms or configurations, or if above binaries fail to run on your computer, you should download source code and compile it.
Source Code
DSK tool is fully written in C++. | Download |
License
DSK binaries and source code are covered by the Affero GPL version 3 license.